The Economic Consequences of a Catastrophic Earthquake: Proceedings of a Forum (1992)

Since Congress enacted the National Earthquake Hazard Reduction Act in 1977, substantial progress has been made in the area of earthquake hazard identification and strategies to deal with earthquake-related problems. The national program, coordinated by the Federal Emergency Management Agency (FEMA), with participation by the National Science Foundation (NSF), the U.S. Geological Survey (USGS), and the National Institute of Standards and Technology, addresses various issues related to earthquake hazards (e.g., risk assessment, prediction or forecasting, mitigation measures such as improvements of building codes, disaster response, preparedness planning, and recovery).

While many earthquake hazard-reduction issues have been addressed by the national program during the past 14 years, one issue that warrants more attention is the economic consequences of a catastrophic earthquake. Efforts to manage earthquake hazards must include an assessment of the public and private sectors' ability to reduce and recover from earthquake-induced losses. Stricken communities and states should have the ability to maintain sufficient financial stability, thus allowing them to rebuild their economic bases following a catastrophic event.

It is important to recognize that no truly catastrophic earthquake—that is, one that affects production facilities, economic markets, and distribution systems in any significant manner—has occurred in a major population center in the United States since the 1906 San Francisco earthquake. A national concern about potential economic consequences, however, has been heightened in recent years as scientifically based probabilities of future earthquakes in urbanized areas have risen, as their time windows have shortened, and as the social and economic costs of recent earthquakes have increased dramatically.

To address this concern, the National Research Council requested the Committee on Earthquake Engineering (CEE) of the Division of Natural Hazard Mitigation to organize a forum on issues related to potential economic consequences—locally, regionally, and nationally—that could occur following a catastrophic earthquake in the United States. Support for this forum was provided by the Federal Insurance Administration of the FEMA and by the National Committee on Property Insurance.

An Advisory Group for the Forum on Earthquake Economic Issues, chaired by CEE member Dr. Joanne Nigg, was established to identify key issues to be addressed during the forum. The forum, held on August 1–2, 1990, in Washington, D.C., consisted of 15 invited presentations that reflect the state of the art of economic theory, economic modeling, hazard characterization, and societal impacts as they can be applied to concerns about earthquake consequences. The first day and a half of the forum consisted of

these presentations and short discussion periods to clarify points made in the presentations. The last half day of the forum was devoted to a general discussion, in which both the presenters and the audience participated, of issues raised during the meeting.

The following topics were addressed:

1. What is known about the seismic risk nationally?

2. What are the likely categories of loss and damage in the event of a catastrophic earthquake?

3. What is the current state of the art of economic research on earthquake consequences?

4. To what extent does the location of the catastrophic event affect the types and extent of impacts?

5. To what extent can regional and interregional shifts in resources take place following a catastrophic earthquake?

6. How likely is a ''ripple effect'' to occur following a catastrophic earthquake, and what consequences may result from such a phenomenon?

7. How do current state and federal postearthquake policies and programs affect economic recovery capabilities of the public and private sectors?

Regardless of their specific topic areas, presenters were asked to keep the following three orienting questions in mind. These questions were also used to focus the general discussion on the second day of the forum.

1. In the event of a catastrophic earthquake in the United States, what are the potential economic losses within the impact area?

2. Under what conditions might local or regional earthquake losses result in national economic disruption?

3. What conditions, programs, or policies are capable of either exacerbating or reducing the amount of both local and national disruption that could take place?

The invited presentations, including the clarifying questions asked by audience members, are in the second major section of this report. The remainder of this executive summary summarizes the general discussion on the second day of the forum with respect to (1) the major themes identified and the range of views expressed on these themes and (2) the types of research that are needed to resolve problems associated with the thematic issues.

The forum was not intended to provide recommendations on policy matters or needed program initiatives. Its primary objective was to review the major issues related to economic impacts from large-scale earthquakes, to assess current capabilities—in terms of data bases and economic models—to predict these impacts, and to review current state and federal programs related to recovery from such events.

Five thematic issues were extensively discussed during the second day of the forum: (1) the capabilities of and requirements for economic modeling, (2) the nature and scope of possible social and economic impacts, (3) the likelihood of a national "ripple effect," (4) the relationship between insurance and mitigation, and (5) the need for a new approach to the problem.

Panelists and audience members identified the need to project various types of economic impacts from a catastrophic earthquake, including: direct economic losses due to destroyed or severely damaged buildings and other structures (such as dams and lifeline systems), direct economic losses due to damaged or destroyed contents of buildings and other private property, indirect economic losses due to disruption of the private sector (that is, business interruption), loss of revenues and increases in expenses for the public sector, and losses of individual and household income due to injury, death, or job interruption.

Reasonably sound loss-estimation methods currently exist to project direct damage to buildings from ground motion. However, two major limitations of these methodologies were identified. First, the models have been primarily developed for application to engineered buildings and are less reliable when being applied to nonengineered buildings or to other types of structures (for example, to lifeline systems). Estimates of loss, therefore, are better for large, multistory, commercial and residential buildings than for smaller or, often, nonengineered (one- or two-story) buildings or residential dwellings.

Second, currently available data bases required by these models to develop reliable loss estimates at the community level are inadequate on a national basis. This is especially true outside of California. The primary type of information that is unavailable is an inventory of existing buildings and structures. Such an inventory might include: ages of the structures, construction types (e.g., configuration, height, and materials), usage patterns, and number of inhabitants. From this structural inventory, response to ground motion estimates, and hazard characterizations, projected damage estimates could be made, including economic costs for replacement and repair of the structures, following earthquakes of various magnitudes for specific communities. Without good inventory data, however, these damage estimates are, at best, educated guesses.

The forum discussions illustrated that currently no one methodology is generally accepted for projecting likely damage to the contents of structures or to other types of private property that are likely to be affected by structural damage. The magnitude of the losses that could be associated with damage to production machinery, office equipment (including computers), inventories, or raw materials due to collapsed or partially collapsed structures is unknown. The indirect consequences of these types of losses (in terms of business disruption) are discussed below. However, damage or destruction of these resources has an immediate, direct economic impact. That is, equipment and supplies must be replaced before production can resume.

Two kinds of problems were raised concerning damage projections. First, models have not been developed that relate direct structural damage to nonstructural, content losses. Second, such models would have to rely on inventories that were even more extensive than those required to project structural losses. Although some modeling work has been undertaken to project nonstructural (or content) losses to a specific building, the work does not extend to classes of structures.

Business interruption can result from damage to the structure in which the business is located; damage to production or manufacturing equipment, office equipment, and inventory, loss of production materials due to losses experienced by a supplier whose facilities were also damaged; loss of electrical power or other lifeline services necessary to operate the facility; interruption of the transportation system to deliver supplies or finished products; loss of customers due to damage to their facilities or inability to access the facility; or loss of employees due to death or injury.

Little is empirically known about disaster-generated indirect economic effects on business in general, and less on the consequences for specific economic sectors. However, recent postearthquake research has indicated that few small businesses affected by an earthquake are likely to have business interruption insurance. Several participants stressed that this situation could result in an increase in the business failure rate following a large-magnitude earthquake.

The participants reported that some economic research has been undertaken on market interdependencies at a national level, but little effort appears to have been expended on this problem at a subnational level. These participants also indicated that to begin to sufficiently address this concern at either a regional or interregional level, data bases must be developed on the geographic distribution of trade flows to make the available models operational.

While many participants felt that such economic loss projections would be useful, there was little agreement on how accurate such estimates had to be to provide policy advice to decision makers in the public or private sectors.

A catastrophic earthquake could affect government at all levels, but especially at the local and state levels, by reducing future revenues, increasing current costs resulting from response activities, and increasing future costs resulting from recovery and reconstruction activities. However, there could also be revenue increases to offset these decreases: for example, construction permits and new business licenses (for construction-related businesses), and temporary occupancy fees (bed taxes) paid by temporary repair and relief personnel. Nevertheless, the extent to which these increases and decreases would be balanced, and for which levels of government, is yet to be explored.

The participants reported that no systematic research has been conducted on the overall economic effects of a major disaster on the public sector, much less on trying to project these impacts for a future catastrophic earthquake; nor do models or data bases currently exist to estimate these different types of economic costs to government. Also, despite the fact that emergency service professionals have expended a great deal of effort on emergency-response planning for a destructive earthquake during the past two decades, these plans have never been used to estimate governmental budgetary needs. These increased postearthquake governmental costs, as described at the forum, include: debris removal and disposal, urban search and rescue efforts, fire-fighting and hazardous-materials-event response, provision of emergency medical services, provision of temporary shelter, overtime for salaried governmental workers to perform a variety of operational and administrative services, and inability to invest in new, productive projects because of repair and reconstruction costs associated with publicly owned damaged facilities.

Similarly, the participants noted that no methodology exists for estimating the impact that a large-scale disaster could have on government revenues. Revenue losses could result from changes in property tax assessments (due to structural damage), reduction in both business and personal income taxes, declines in purchases resulting in lower sales taxes, declines in bed taxes due to a decline in tourism, and fewer user fees, among others.

Another unknown factor, as reported at the forum, is the extent to which the ability of a community to issue municipal bonds to fund government-sponsored development projects would be affected by a large earthquake. Some participants stressed the importance of this issue, since communities may have to pass additional bond measures to fund reconstruction projects for government-owned buildings and facilities. If currently funded projects are significantly damaged, resulting in the need to expend additional public monies to complete the project, postearthquake reconstruction projects may be further

delayed because of inability to raise capital (e.g., through the sale of local bonds).

Some presenters indicated that, following a federally declared disaster, households are generally able to recover; but recovery is affected by the nature of the losses, the degree of assistance provided, and the preearthquake socioeconomic status of the household and the community.

Insurance industry representatives at the forum expressed concern about the consequences to the industry if the need arises to cover a wide variety of personal and household loss claims from a catastrophic earthquake. In addition to shake damage, which is covered by an earthquake insurance policy, there may be additional indirect losses, such as workers' compensation, medical costs, life insurance, liability coverage, and automobile damage. Such a catastrophic loss could erode insurers' surpluses, potentially resulting in an availability crisis (i.e., a market failure). The participants reported that while some information is available to estimate the extent of "covered" losses (i.e., covered by the insurance industry), no information currently exists to project all losses (including those that are not covered by insurance) for these categories.

The participants noted that almost no solid empirical data are currently available regarding the nature, scope, and duration of likely economic and social impacts that would result from a catastrophic earthquake. They further identified a definite need to better understand these effects on several dimensions: direct and indirect losses and costs, economic and social losses and costs, impacts for various governmental levels (local, substate, state, regional, and national), and the consequences of disruption and dislocation for households, businesses, and communities.

In addition to the specific modeling issues addressed above, many participants expressed the belief that net economic impacts of catastrophic events are likely to be negative, at least for a short time at the substate level; but the likely duration and distribution of these impacts, especially for various economic sectors (e.g., manufacturing and retailing) and for specific social groups is largely unknown.

One major issue surfaced—the need to address the redistribution effects and consequences of large earthquakes. Even if impacts are not felt beyond the local or substate area, models need to be developed and attention paid to the "winners" and "losers" of these redistribution effects between, as well as within, economic or business sectors. Similarly, ethnic and class groups will be differentially affected by a catastrophic earthquake, resulting in their

differential ability to recover. While aggregate changes in the economic health of an area may be only temporary, sectoral changes may have longer-term consequences for the area's labor force, minority-owned businesses, or neighborhood property values. Such changes, as reported by some participants, could substantially affect the quality of life and the economic viability of the community. Mitigation and recovery planning must take these qualitative consequences, as well as the economic ones, into account.

Extensive discussion focused on the likelihood that a catastrophic earthquake could generate widespread, unacceptable, negative economic consequences for the insurance industry and, by extension, for both the public and private sectors.

In part, the differences of opinion expressed during this discussion were philosophic. Those who maintained that economic consequences were more likely to be minor, local, and time-bound referred to the resilience of the market even during the worst economic periods in recent U.S. history (e.g., the stock market "crash" of October 1987). Those who painted a "doomsday" scenario argued that financial markets are extremely sensitive to any uncertain situation. For either argument, the participants agreed that with respect to large-scale natural disasters in recent history, there is very little empirical evidence available, other than data on immediate, short-term impacts, upon which to test either of these points of view. Some participants indicated that they do not believe a catastrophic earthquake would have a crippling effect on the national economy of the United States, and many stated that they felt the scope and duration of economic consequences could vary, depending on where such an earthquake occurs. Outside California (for example, in the central United States), greater direct structural and lifeline system damage would be expected (since a larger proportion of the existing building stock and lifeline system was constructed without regard for seismic design), resulting in proportionally greater economic losses and social consequences.

Because of the insurance industry's stated concerns about the erosion of their surplus capital as a consequence of a catastrophic earthquake, some of the forum discussion focused on the feasibility of earthquake insurance as a mitigation tool. Many participants believe that earthquake insurance, as it is currently offered, is not an efficient mechanism to reduce economic impacts on homeowners and businesses, for several reasons. In the highest-risk areas in California, for example, premiums are very high and deductibles very large, which has discouraged greater earthquake-insurance purchases. In high-risk areas outside of California, premiums may be substantially lower, but there is

also a lower demand for earthquake insurance by the public, either because people do not perceive the risk to be sufficiently large to require coverage or because they are under the mistaken impression that their general homeowner's policy will cover earthquake-caused damage. The participants are in general agreement that the cost of inspecting residential structures to determine vulnerability or to verify levels of mitigation is prohibitive today, relative to the amount of premiums charged. Current insurance practices, therefore, do not tie rates to mitigation.

Much discussion focused on the need to develop a federal earthquake reinsurance fund to increase the insurance industry's ability to avoid a market failure in the event that a catastrophic earthquake occurs in a major metropolitan area. There was disagreement on the need for such a program, because it is unclear whether the insurance industry needs federal assistance to cover earthquake-related losses and because there is general skepticism about a national ripple effect. Nonetheless, participants on both sides of the issue agreed that if such a federal program were to be developed, the rate structure should be risk-based. That is, it should reflect the actual risk to the structure. Both sides also agreed that the rate structure should reward, through lower premiums, individual and community mitigation efforts to reduce that risk.

Throughout the forum discussions, it was apparent that the problem of anticipating the economic impacts and consequences from a catastrophic earthquake is not a problem that should be addressed either in a piecemeal fashion or by a single discipline or industrial sector. The participants nevertheless agreed that, unfortunately, much of the work done to date has suffered from one or both of these deficiencies. (One notable recent exception, however, is the FEMA-sponsored project, Loss-Reduction Provisions of a Federal Earthquake-Insurance Program.)

The participants further agreed that future research efforts should be multidisciplinary, involving university as well as public and private sector research teams. Because there was such a great emphasis among forum participants on the utility of these projects for government mitigation programs and fiscal planning, these research projects, as indicated by the participants, should also have multidisciplinary public- and private sector advisory panels to assist in the refinement of research objectives and to provide ongoing input to the projects, including the provision of data for use in the economic models.

Four types of research needs were identified by the forum participants: loss-estimation models, economic models, redistributional effects, and policy issues.

1. Loss-estimation models would be improved if they could project direct losses to structures other than buildings, contents of buildings, possible fires following earthquakes, and the factors used in econometric models to describe economic change.

2. Applicability of the models will require inventories of structures and their contents. A methodology for developing such inventories could be formulated to make these models work.

3. Loss-estimation models will also require more-detailed characterization of the hazard, including projections of ground motion, from earthquakes of various magnitudes in high-risk areas.

4. Economic models and data are needed that can accurately project how a catastrophic earthquake would disrupt the flow of trade and commerce at regional, state, and substate levels.

5. These models should be developed based on the potential impacts of the critical "engines" that drive the economies in the various earthquakerisk regions of the country.

6. Methodologies are needed that can project likely economic redistributional effects both between and within states.

7. Further, these methodologies should be refined to examine the differential redistributional effects across social groups and communities in possible impact areas to identify causes and, if necessary, solutions to negative impacts.

8. A better understanding of the social and economic consequences resulting from different mixes of mitigation and hazard-reduction policies would allow local and state governments to evaluate issues such as what are

the benefits? who benefits and who loses? what are the costs? who pays these costs? what are the social effects of implementation? and how much public finance is justifiable?

9. An analysis is needed of the existing state and federal aid policies to assess their adequacy and consequences, especially for local governments.

10. Economic models that would provide government decision makers useful information for formulating mitigation and economic recovery policies can be formulated. These decision makers should be closely consulted in developing the needed level of precision of these models for use in their respective communities or regions.