CRITICAL REVIEW OF METHODOLOGIES ON DISASTER IMPACT

1

CRITICAL REVIEW OF METHODOLOGIES ON DISASTER IMPACT

ESTIMATION1

YASUHIDE OKUYAMA

Graduate School of International Relations, International University of Japan, Niigata,

Japan

ABSTRACT

The main aim of this paper is to provide an overview and a critical analysis of the

methodologies used for estimating the economic impact of disaster. Based on the

critical review of methodologies, especially those recently progressed, such as

input-output, social accounting, and computable general equilibrium models, the paper

presents the strengths and weaknesses of the methodologies. In addition, the issues of

impact estimation when applied to developing countries are discussed and analyzed.

1 This background paper was prepared as an input into the joint World Bank – UN Assessment on the

Economics of Disaster Risk Reduction. The author would like to express gratitude to Apurva Sanghi

and Sebnem Sahin for their support, encouragement, and patience.

2

1. Introduction

Until the 1980s, the economic impacts of natural hazards and disasters had received

relatively little attention from both researchers and practitioners. Recent large-scale

disasters, such as the 2004 Indian Ocean Earthquake leading to the Asian Tsunami and

Hurricane Katrina in 2005, and more recently the 2008 Cyclone Nargis in Myanmar and

the 2008 Sichuan Earthquake in China, showed that the understanding of and the

preparation for such hazards, including their impacts and effects to society and the

economy, are urgent tasks around the globe. In addition, a series of disasters in the

mid 1990s, such as the Northridge Earthquake in 1994 and the Kobe Earthquake in

1995, which occurred in developed urban areas and brought considerable damages and

impacts to the society, demonstrated how vulnerable even modern societies are to severe

natural hazards.

While significant progress has been made in recent years in the economic

analysis of disasters, especially in the field of economic modeling for disaster impact

(for example, Okuyama and Chang, 2004a; and an excellent compilation of related

papers by Kunreuther and Rose, 2004a), these have been mostly in the developed

country context. Since the pioneering work by Dacy and Kunreuther (1969), a

generalized framework for the economic analysis of natural disasters had been proposed

(for example, Sorkin, 1982; and Albala-Bertrand, 1993a). The recent advancements

have been more toward empirical analysis and toward strategies for modeling

extensions and modifications to fit them to disaster situations. This trend is due to

improved data availability of disaster damages and losses and to increased

multidisciplinary research activities about disasters, including sociology, economics,

and psychology. The uniqueness of each hazard and of its damages and impacts

presents enormous challenges in economic modeling for disaster impact estimates;

many issues remain unsolved, especially in the developing country context. For

example, while the domination of flow analysis models employed in the above research

indicates that in developed countries the damages to capital stock are rather minor in

terms of macroeconomic context, the damages to both physical and human capital

stocks by a catastrophic disaster in developing countries can become truly devastating,

3

potentially washing out a major portion of capital stock in a nation, such as the 1998

floods in Bangladesh (Shah, 1999). In this paper, issues related to disaster impact

estimation in the developing country context are raised and discussed, and various

methodologies for doing so are also reviewed.

It is helpful to first clarify the concept and definition of disaster, since unclear

terminology of the event has caused some confusion about what the impact estimate

should cover. Several terms, such as disaster, hazard, unscheduled event, catastrophic

event, among others, have been used interchangeably in the literature; however, not all

disasters or hazards lead to catastrophic consequences, nor are all hazards or disasters

unscheduled events. In this context, the two terms, ―disaster‖ and ―hazard‖, include a

wider range of events than the others. The distinction between disaster and hazard can

be found in Okuyama and Chang (2004b, p. 2); ―hazard is the occurrence of the

physical event per se, and disaster is its consequence.‖ Ariyabandu (2001) put this

more specifically suggesting that a disaster is an outcome of a hazard impacting on the

vulnerability of a society. Thus, the present paper uses the term, disaster, as the event

that the methodologies are employed to estimate its impact.

In the following section, economic modeling practice for disaster impact

estimate is critically reviewed. Section 3 discusses the current issues of modeling,

such as time, space, and the counteractions. Section 4 discusses the disaster impact

assessment in the developing country context, focusing particularly on long run impact

and negative externalities. The final section concludes with some future directions for

promoting and improving the methodologies.

2. Methodologies on Disaster Impact Estimation: Models for Impact on Flows

Natural disasters can cause physical destructions to built-environment and networks,

such as transportation and lifelines, and can also cause casualties and injuries to human

lives. These damages are often called damages, and are by economics definition the

damages on stocks, which include physical and human capitals. These damages lead

to the interruptions of economic activities, such as production and/or consumption, and

the losses from business interruptions are often called the losses of a disaster. At the

4

same time, there is another term called indirect (or secondary) effects, which take into

account the system-wide impact of (indirect) flow losses through interindustry

relationships. For example, if a power company‘s production facility is damaged by a

natural hazard, the power company‘s business is interrupted and cannot produce

electricity for some time until repaired. This production loss is the loss of this

particular power company. On the other hand, another firm, which uses electricity for

their offices, is not damaged by the hazard but cannot operate their business fully

without the supply of electricity, and needs to reduce or stop their production until

power is restored. The business disruption of this firm is called the indirect effect. In

this sense, indirect effects are the system-wide impact of losses.

While some researchers critique that the indirect effects of disaster are ―more a

possibility than a reality‖ (Albala-Bertrand, 1993a, p. 104), the estimation of indirect

effects has been attempted to ―gauge individual and community vulnerability, evaluate

the worthiness of mitigation, determine the appropriate level of disaster assistance,

improve recovery decisions, and inform insurers of their potential liability‖ (Rose, 2004,

p. 13). Indirect effects include a wide array of consequences caused by damages, and

Rose (2004) suggested that the use of ‗higher-order effects‘, instead of indirect effects,

is more appropriate, because of the conflict with the terminology used in economic

models, especially with input-output (IO) model. In this context, losses can be

considered as first-order effects from damages. In contrast to damages on stocks,

losses and higher-order effects are mostly on flows, such as services or output of stocks

over time. And with the use of higher-order effects, another term, total impacts which

are the total of flow impacts, adding losses (first-order losses) and higher-order effects,

should be introduced.

Various economic modeling frameworks have been employed to estimate the

higher-order effects of a disaster. Perhaps, the most widely used modeling framework

is the IO model (for example, Cochrane, 1974, 1997; Wilson, 1982; Kawashima et al.

1991; Boisvert, 1992; Gordon and Richardson, 1996; Rose et al. 1997; Rose and

Benavides, 1998; and Okuyama et al., 1999), and the application of the IO model to

disasters, including both natural and man-made ones, dates back to strategic bombing

studies during World War II (Rose, 2004). The popularity of IO models for disaster

related research is based mainly on the ability to reflect the economic interdependencies

5

within a regional economy in detail for deriving higher-order effects, and partly on its

simplicity. The simplicity of the IO framework has enabled integrative approaches, in

which IO models are combined with engineering models and/or data, in order to

estimate higher-order effects that are more sensitive to the changes in physical

destruction. Some examples of this approach include the links with transportation

network models (Gordon et al., 1998, 2004; Cho et al, 2001; Sohn et al., 2004, among

others), with lifeline network models (Rose, 1981; Rose et al. 1997; Rose and

Benavides, 1998), and the comprehensive disaster assessment model, namely HAZUS

(Cochrane et al., 1997).

On the other hand, this simplicity of the IO model creates a set of weaknesses,

including its linearity, its rigid structure with respect to input and import substitutions, a

lack of explicit resource constraints, and a lack of responses to price changes (Rose,

2004). In order to overcome these weaknesses in the context of disasters, several

attempts of refinement and extension of the IO framework have been proposed. For

instance, the shortage of regionally produced inputs under a disaster situation was dealt

with by the integration of a methodology for more flexible treatment of imports

(Boisvert, 1992; and Cochrane, 1997). The issue of supply-side constraints due to the

damages to production facilities was addressed with the allocation model variant of IO

model (Davis and Salkin, 1984); however, this modeling scheme has inherent

deficiencies (Oosterhaven, 1988 and 1989; see Dietzenbacher, 1997, for a solution), and

was later modified by Steinback (2004) to include only backward-linkage effects. The

treatment of price has been transferred to computable general equilibrium (CGE)

models, and the applications to disaster situations are discussed below.

An alternative modeling framework to the IO model is CGE analysis (for

example, Boisvert, 1992; Brookshire and McKee, 1992; Rose and Guha, 2004; and

Rose and Liao, 2005). Unlike IO models, CGE models are non-linear in common

practice, can respond to price changes, can incorporate input and import substitutions,

and can explicitly handle supply constraints. However, Rose and Liao (2005) claim

that most CGE models are intended for long-run equilibrium analysis; hence, in contrast

with the rigidity of the IO model, a CGE model generally leads to the underestimation

of economic impacts due to its flexible adjustment feature. It is argued, however, that

CGE models provide lower impact estimates than IO models, partly because ―not all

6

causations in CGE models are unidirectional, i.e., functional relationships often offset

each other‖ (Rose, 2004, p. 27). Another weakness of CGE models is that the

assumption of optimizing behavior can be considered questionable under disaster

situations, where increased uncertainties arise in the near and distant future. Of course,

the more extensive data requirement for CGE modeling presents a major weakness for

empirical analysis of disasters.

Other modeling frameworks have been also employed to estimate higher-order

effects of disasters. The social accounting matrix (SAM) has been utilized to examine

the higher-order effects across different socio-economic agents, activities, and factors.

Notable studies using a SAM or one of its variants include Cole (1995, 1998, and 2004)

among others. Like IO models, the SAM approach has rigid coefficients and it tends

to provide upper bounds for the estimates. On the other hand, the SAM framework, as

well as extended IO models2 and CGE models, can derive the distributional impacts of

a disaster in order to evaluate equity considerations for public policies against disasters.

Econometric models, which are based on time-series data that may not include any

major disasters, appear ill-suited for disaster impact analysis. Ellson et al. (1984),

however, argue that they examined the damage estimates of major earthquakes in the

United States and found that the damages do not appear really outside of the historical

variability of the regional economy in response to more traditional shocks and cyclical

fluctuations. Furthermore, econometric models are statistically rigorous, can provide

stochastic estimates, and have forecasting capabilities. They do require a large data set

(time-series as well as cross-section) though, and cannot easily distinguish between

direct and higher-order effects (Rose, 2004). The strengths and weaknesses of these

four approaches are summarized in Table 1:

Table 1. Summary of Methodologies

Strengths Weaknesses

IO - simple structure

- detailed interindustry linkages

- wide range of analytical

techniques available

- linear structure

- rigid coefficients

- no supply capacity constraint

- no response to price change

2 The disaster related studies using extended I-O model include Okuyama et al. (1999).

7

- easily modified and integrated

with other models

- overestimation of impact

SAM - more detailed interdependency

among activities, factors, and

institutions

- wide range of analytical

techniques available

- used widely for development

studies

- linear structure

- rigid coefficients

- no supply capacity constraint

- no response to price change

- data requirement

- overestimation of impact

CGE - non-linear structure

- able to respond to price change

- able to cooperate with

substitution

- able to handle supply capacity

constraint

- too flexible to handle changes

- data requirement and calibration

- optimization behavior under

disaster

- underestimation of impact

Econometric - statistically rigorous

- stochastic estimate

- able to forecast over time

- data requirement (time series

and cross section)

- total impact rather than direct

and higher-order impacts

distinguished

Most of the above methodologies3 reviewed are considered as flow analysis

models estimating the higher-order effects of a disaster, which are the effects on flows.

The reason for the popularity of flow analysis models in estimating disaster impacts,

according to Rose (2004), is that ―flow measures are superior to stock measures in many

ways‖ (Rose, 2004, p.14). The strengths of flow measures can be summarized as: 1) it

can measure the impacts (business interruptions) without stock damages; 2) it is a

3 While the study by Ellson et al. (1984) employed the econometric model with the mixture of flow and

stock variables, econometric model can be tailored to include only stock or flow variables, depending on

the research interest. Some examples of econometric analysis with only stock variables are discussed in

the following section.

8

performance measure whereas stock measure involves the life-cycle assessment of

capital with depreciation; 3) it is more consistent with other conventional

macroeconomic indices, such as Gross Domestic (Regional) Product (GDP or GRP);

and 4) it shows the short-run impact of a disaster, which is oftentimes convenient for

policy discussions against disasters.

In addition, the above models have been used for both ex-ante and ex-post

analyses of disasters. Ex-ante analysis of disasters can be done with a series of

hypothetical scenario runs and/or by integrating Monte-Carlo simulation with the model

to produce probabilistic impact estimates (for example, Shinozuka and Chang, 2004,

and Shumuta, 2004). While these simulations can provide policy implications and

insights for the near future events, more stochastic forecast toward the distant future

should rely on econometric models. With the econometric models, the impacts of

future disasters are turned to become social and economic risks with the probability, and

the feasibility analysis of the countermeasures against disasters over time can utilize the

results (for example, Freeman et al., 2002, and Cardenas et al., 2007). For the ex-post

analysis of disasters, the above methodologies are used to a wide range of disasters, as

reviewed above.

3. Methodological Issues

Features specific to disasters, such as negative (destruction) and the following positive

(recovery and reconstruction) shocks to an economy in a short period of time, a wide

range of physical damages across multiple locations, and behavioral changes in a crisis

situation, should be incorporated to the methodology in order to clearly articulate what

is going on in the aftermath of a disaster. However, most economic models assume

gradual and incremental changes over time and uniform or monotonic effect over space.

Thus, when applying such models, it is difficult to cope with the features specific to

disasters. In the following, attempts to extend the modeling framework to handle these

issues are discussed.

3.1. Time

9

The duration of a hazard occurrence, not of a disaster (see above for the difference in

their definitions), varies from one hazard to another: ground shakes of an earthquake

last around 30 seconds to one minute, while a large-scale flood can remain from a few

weeks to a few months. However, these time frames are still considered short in an

economic modeling sense. Mainly due to data availability and to the

equilibrium-oriented assumption, the time frame of most economic models is annual

while some models operate with a quarterly or monthly frequency. Moreover, the

activities of recovery and reconstruction after a hazard are conducted much more

rapidly and simultaneously in multiple locations than usual construction activities.

These features of a disaster generate seesawing temporal changes and lead to

considerable temporally-intense interactions of economic activities. Hence, the use of

static economic models potentially cancels out the positive and negative impacts in the

long run and then often ends up estimating insignificant total impacts (Albala-Bertrand,

1993).

Attempts to incorporate the dynamic nature of disaster situations have indeed

been made. Cole (1988, 1989) proposed lagged expenditure models through extending

the standard IO framework. Concerning the time necessary to produce goods and

taking into account the process of labor market adjustment, among other things, Cole‘s

lagged IO models aim to capture the process of impact (or growth) from a factory

closure (or production expansion) within the IO framework. With a similar objective,

but putting more emphasis on production chronology and production modes, Okuyama

et al. (2004) employed the sequential interindustry model (SIM) for disaster impact

analysis, creating a quarterly IO model with the SIM modification. The SIM

framework was originally introduced by Romanoff and Levine (for example, Romanoff

and Levine, 1986) in response to the need for analyzing interindustry production in a

dynamic economic environment, such as large construction projects where the effects

on production and employment are transitory. Okuyama (2006) later extended the

SIM with the inclusion of an inventory function, and demonstrated the temporal

distribution of higher-order effects from hypothetical lifeline disruptions.

Although the SIM framework can handle the shorter time period better than the

conventional economic models do, in order to provide greater insight into temporal

interindustry linkages, it is still not sufficiently flexible because of its discrete nature.

10

For a more flexible modeling scheme in terms of time, Donaghy et al. (2007) proposed

a continuous-time formulation of a regional econometric-input-output model (REIM) to

capture higher-order effects of unexpected and extreme events, i.e. disasters, and of

nonlinear continuous event recovery processes. Their continuous-time model

demonstrates the capability to easily accommodate a range of short- and long-term

phenomena in disaster situations. Another direction may be the application of

dynamic CGE models; however, the issue of temporal substitution in a disaster situation

can become troublesome, and its application has not been made to disaster situations to

this date. While these extensions of the conventional economic models shed some

light on the temporal feature of disaster impacts, the modeling scheme is still considered

somewhat ad hoc, since no definitive theories have been established for their application

to the idiosyncrasy of conventional models (Rose, 2004).

3.2. Geographical Space

Since disasters can create a wide range of impacts over space, the spatial dimension of

higher-order effects has been dealt in many ways. One way to do this is the

development of an interregional impact analysis of a disaster, because a large natural

hazard can create impacts that spread well beyond the boundary of the region in which

the hazard occurred. For example, Okuyama et al. (1999) employed a two-region

interregional IO table to estimate the higher-order effects of the 1995 Kobe Earthquake,

both in the region where the earthquake occurred and in the rest of Japan. As

mentioned above, various integrative approaches with transportation models have been

attempted for analyzing the spatial distribution of higher-order effects. For instance,

Sohn et al. (2004) calculated the multi-regional impact of a hypothetical New Madrid

Earthquake in the United States, utilizing the interregional commodity flow model

based on the Leontief-Strout-Wilson-type modeling framework. However, these

studies are based on the IO framework, and the changes in interregional trade patterns

are dealt with either in an ad hoc way or in a transportation engineering way

(minimizing the total transportation time). In this line of research, Tsuchiya et al.

(2007) proposed a spatial CGE (SCGE) model, which includes a transportation model

that can estimate interregional freight flows and passenger trips, to measure the impacts

of transportation network disruptions by catastrophic events. This approach is

11

particularly useful when the price difference of some commodities between regions

takes place due to damages in one region‘s economy and/or transportation disruptions

between regions, causing the increase of transportation cost. Their SCGE model also

includes the feature of short-run equilibrium, rather than long-run equilibrium as in

usual CGE analysis, to accommodate with the conditions under a disaster situation,

including the restriction on the movements of labor and capital.

The damages and losses of a disaster, such as a large earthquake or a tornado,

are often not uniform across the region: some areas are heavily destructed, whereas

other areas are virtually untouched. In order to analyze the sensitivity of spatial

distribution of damages and higher-order effects, the following extensions have been

made using spatially disaggregated economic models. Cole‘s (1998) multi-county

SAM was constructed based on county-level economic data and geographic information

system (GIS) based location data, and was applied to lifeline failures in the Memphis

region. Disaggregating further economic information to smaller geographical areas,

van der Veen and Logtmeijer (2003) employed the concept of economic hot spots,

through which multipliers from the IO table are visualized as a contour map using GIS.

This visualization does not directly provide the detailed information about the

higher-order effects of a disaster, but can illustrate which locations are more vulnerable

(or important) in terms of economic interdependency. Extending the concept of

economic hot spots, Yamano et al. (2007) proposed a disaggregation of a multiregional

IO model to a district level (500-meter grid) in order to show the spatial distributions of

economic activities and of potential higher-order effects. The strength of their finer

geographical-scale model is to clearly identify which district is crucial in terms of

economic losses.

3.3 In-Built Counteractions

As mentioned before, Albala-Bertrand (1993, p. 104) claims that indirect (higher-order)

disaster effects are ―often unimportant for the economy and society as a whole and

rapidly counteracted within the disaster area,‖ since ―in-built societal mechanisms may

prove sufficient to prevent most potential indirect (higher-order) effects on the economy

and society.‖ Meanwhile Dacy and Kunreuther (1969, p. 64) indicated the sympathetic

behavior of mutual aid in a chaotic situation, leading to the situation that ―the supply

12

and demand curves may shift unexpected ways‖ in extreme circumstances. These

counteractive mechanisms and behavioral changes have started to receive increased

attention in recent years (Kunreuther and Rose, 2004b). Some attempts to incorporate

behavioral changes, as negative externality, in a disaster situation have been made.

For example, Okuyama et al. (1999) included the final demand decrease in the rest of

Japan after the Kobe earthquake, since people outside of the damaged region felt sorry

for the event and for people in the Kobe area, due to the catastrophic destruction of a

major city and a large number of casualties, and tended to reduce their discretionary

purchases. Or, people may purchase necessary goods for the damaged area, such as

blankets and/or food, and may donate them for helping the people in the damaged areas,

instead of buying other goods for themselves. These types of behavioral changes

(consumption pattern changes) have not been well studied or modeled yet, although

these behavioral changes are a part of the in-built counteractions.

Another type of in-built counteractions is the so-called economic resilience,

―which refers to the inherent ability and adaptive response that enables firms and

regions to avoid maximum potential losses‖ (Rose and Liao, 2005, p. 76). In the

disaster related literature, Tierney (1997) studied the business coping behavior and

community response in disaster situations, Bruneau et al. (2003) introduced community

resilience as the in-built counteractions of a community, and Rose and Liao (2005)

modeled economic resilience successfully in the CGE modeling framework. In the

latter model, inherent resilience, i.e. the ability to substitute inputs and/or reallocate

resources under normal circumstances, is embodied in the production function for

individual businesses, while adaptive resilience, i.e. the ability in crisis situations with

extra effort, is set as the changes in the parameters. While the extent of resiliency

across economic sectors has been studied empirically by Kajitani et al. (2005a and

2005b), the further studies on theoretical foundation of economic resilience need to be

carried out so that more comprehensive impact estimate of a disaster, taking into

account of resiliency mechanism, can be produced.

4. Measurement Issues in the Context of Developing Countries

13

The empirical studies with the above flow analysis models, or other methodologies,

often found that the estimated total impact of a disaster in a short run, i.e. the sum of

negative impacts caused by direct losses and higher-order effects and positive impacts

from recovery and reconstruction activities, become negligible or even positive

(Albala-Bertrand, 1993a and 1993b, Tol and Leek, 1999, Okuyama et al., 1999, among

others). On the other hand, the methodologies reviewed above can cover only the

short run economic consequences and can hardly cover the full scale of impacts,

including the long run effect on economic growth, psychological impacts, deterioration

of public health, personal losses, and livelihood disruptions, which may be

proportionally higher in developing countries (Pelling et al., 2002). In this section, the

issues of disaster impact estimate, especially in the development context, are discussed.

4.1. Development

Most empirical studies with cross-country data investigating the relationship

between development level and disaster losses conclude that correlation between them

is negative, i.e. ―the higher the level of development, the smaller both the number of

deaths, injured, and deprived, and the relative material losses‖ (Albala-Bertrand, 1993a,

p.202). Similar results are found in Anbarci et al., 2005; Kahn, 2005; Toya and

Skidmore, 2005, among others. This appears consistent with the disaster theory that as

countries develop and grow, they should have sufficient resources, such as financial

and/or technological ones, to better manage disaster risk through the implementation of

countermeasures and to better manage the adverse impact of disasters. Some policy

analysts have suggested that a way to reduce disaster damages and losses for less

developed countries should be to develop and grow faster (for example, Okonski, 2004,

and Hoke, 2005).

However, with the recent increasing complexity of social and economic

structure and interdependency within and across countries, this relationship may not be

as straightforward as it seems. According to Lester (2008), direct losses from disaster

(as % of GDP) appear to have a negative correlation with GDP per capita; however, as

GDP per capita increases, the complexity of economic system also increases and thus

the higher-order effects (as % of GDP) has a positive correlation with GDP per capita;

as a result, the total impact over GDP per capita has an inverted ―U‖ shape curve. This

14

implies that the most potentially affected economies by disaster will tend to be

middle-income-level economies. This observation is shared with Benson and Clay

(1998). They claimed that the most vulnerable economies are not the most

underdeveloped, since least developed countries tend to have simple economic

structures, such as agriculture. While middle-income-level economies with some

diversifications seem more secure, because of intertwined economic activities between

industries, however, the higher-order-effects can be much greater than in a simple

agro-economy, and the total impacts from a disaster can be larger than in a simple

economy. This is exactly the point that flow analysis models aim to address—the

complexity of economic activities across industries lead to spread higher-order effects

to larger extent. At the same time, Kellenberg and Mobarak (2008) claim that this

inverted U shape relationship could be attributable to behavioral changes at micro level

in response to income increase. They argue that the risk-return trade off will swing in

favor of return (consumption) at low levels of income where the marginal utility of

consumption is higher, but then will swing to risk mitigation as income increases (the

marginal utility of consumption becomes relatively low). Their empirical analysis of

133 countries over 28 years showed the stronger tendency of this inverted U shape

non-linearity for floods, landslides, and windstorms than for extreme temperature events

or earthquakes. This result implies that ―the achievement of the simultaneous goals of

natural disaster risk reduction and poverty elimination cannot be assumed to be

complementary for all disaster types‖ (p. 779). In this context, particularly for ex-ante

analysis to assess disaster risk, it is advisable to control income level and disaster type

in the analysis of disaster impact so that different effects are taken into account.

4.2. Long Run Growth

Only a limited number of empirical studies4, compared to the number of studies

on impact studies or on development, have been carried out for investigating the effects

of disasters on long run economic growth caused by stock losses. Skidmore and Toya

(2002) found that while climatic disasters are positively correlated with long-term

economic growth, investment on human capital, and total factor productivity growth,

4 The recovery from war destructions has been studied slightly more often (see for example, Davis and

Weinstein, 2002 and 2004; and Miguel and Roland, 2005).

15

geological disasters are negatively correlated (or sometimes statistically insignificant)

with economic growth. They attributed this result to the notion that climatic disasters

(frequent and predictable to some extent) are threat to property but not life, whereas

geological disasters (less frequent and unpredictable) are threat to both property and life.

Moreover, their results also indicated that climatic disasters are associated with the

growth of total factor productivity, indicating that climatic disasters may provide the

opportunity to update capital stock and to adapt new technologies. This last point is

the motivation of the research by Cuaresma et al. (2008), in which the relationship

between technological transfer to and disaster in developing countries in the long run is

examined. Their results contradict with Skidmore and Toya‘s and instead find that

disaster risk is negatively correlated to the extent of technological transfer, while only

countries with higher level of development (higher level of per capita income) can

benefit from technological transfer after a disaster. Similar results to Cuaresma et al.

are found in Rasmussen (2004) with the data of countries in the Eastern Caribbean

Currency Union, which indicate that the long run effects of natural disasters on growth

are inconclusive. This disagreement of the results among the above studies may be

due to the fact that ―disaster variables are somewhat crude measures‖ (Skidmore and

Toya, 2002, p. 682), and especially for the intensity of disaster (physical and economic

losses) variable, there is no standard definition or method devised and applied across the

cases. This issue has to be addressed in order to promote and encourage this line of

research.

Lack of research in this area may have resulted from several other reasons as

well: the (in)availability of reliable data on capital losses, especially on human capital;

the complexity of assessing the value of capital over time; the intricacy of investment

decisions on damaged capital under and after a disaster (Okuyama 2003); the lack of

policy maker‘s interest in the distant future5; and the difficulty of separating out the

disaster-induced effects in a long run, especially for developing countries where the

influences to their macroeconomic conditions from other factors, such as political

instability, debt burden, among other things, appear much greater than in developed

countries. In addition, since a large portion of empirical case studies employ flow

5 Chambers (1989) stated that the demands of poverty usually out-compete the demands of vulnerability

in developing countries.

16

analysis models for the immediate impact estimate, and these results suggest that most

total effects on economic aggregates become negligible (or not so spectacular) within a

few years (Albala-Bertrand, 1993a), or so assumed (see for example, Okuyama et al.,

1999). However, Albala-Bertrand (1993a) is cautious on this, especially for

developing countries‘ case: ―Nonetheless, there appear to be some important economic

effects which go beyond simple physical reconstruction and improvement‖ (p.200).

Theoretical investigation on this issue also received little interest or attention

from researchers. Dacy and Kunreuther (1969) were first to mention this weakness

and were also first to bring the growth model approach to the analysis of long run

impact. This approach was innovative in that it explicitly included stock damages in

the analytical framework. Okuyama (2003) extended their idea to employ a simple

Solow-Swan model for examining the effects of stock loss by a disaster. Assuming the

economy will return to the same steady state, his analysis indicates increased saving rate

for recovery and reconstruction of damaged capital and slightly slower recovery rate

when the damaged capital is replaced with the one using new and updated technology.

In an economic dynamics and fluctuations context, Hallegatte and Ghil (2008) analyzed

the macroeconomic response to natural disasters using an endogenous business cycle

model6. The analysis of their theoretical model for a hypothetical country, simulating

disequilibria and the adjustment processes, found a ‗vulnerability paradox‘ (p. 586) that

an economy with a high growth rate appears more vulnerable to disasters than with a

slow growth rate (or in a depression) where some production factors are left unused.

With the exception of these above studies, and studies on pollution-induced disasters,

such as global warming and the increase of sea surface temperature (see for example,

Ikefuji and Horii, 2006), the connection between the long run impact of natural disaster

and theoretical modeling approach has not been extended any farther yet. The reasons

for this absence of interest may be similar to the empirical studies‘ discussed above, or

perhaps the theoretical formalization of disaster per se has not reached the level where

the theoretical model can incorporate and handle various complexities (Okuyama, 2007).

This situation, together with the limited empirical studies, needs to be improved in order

6 Business cycle models are often categorized as a short run model in macroeconomics, but in this paper

they are put in this subsection on long run growth because of its longer time horizon than flow analysis

models.

17

to link disasters with development theory (otherwise, it should be proven that disasters

have no long run impact).

In the context of long run growth and disasters, the studies on the impact of

human capital loss have been also very limited. While the publicity of human toll,

such as deaths, injuries, and missing, is usually very high in disasters, these human

capital losses are usually counted as the decrease of labor and/or consumers in flow

analysis models (see for example, Okuyama et al., 1999) or are left out in short run

analysis (Albala-Bertrand, 1993b), because the effect of human capital losses has long

run implications and cannot be fully incorporated in short run analysis. In general,

human capital is much more difficult to ―recover‖ or ―reconstruct‖ than physical capital

per se, because its formation requires education and/or training; health investments and

so on; and the accumulation takes some time. In a disaster situation, this may not be

so dire in developed countries, since there is an adequate size of labor market across the

levels of human capital (from skilled to unskilled) or a sufficient size of idle labor.

However, in developing countries, the situation becomes complicated. Since the

average educational attainment level is usually lower in developing countries than in

developed countries, the quantity of skilled labor tends to be limited. While this

implies that they earn comparatively higher wages than unskilled labor can and thus can

afford to be in a safer and relatively hazard-resistant location and/or structure, on the

other hand, once their lives are lost or damaged (injured), such skilled labor cannot be

easily replaced. Unskilled labor, with low wage level, lives and works in relatively

hazard-prone areas; in fact, most of the casualties in disasters are ―the poorest countries

and their weakest socio-political groups‖ (Albala-Bertrand, 1993a, p.90). Whereas

unskilled labor is at higher risk, such labor appears to be abundant in developing

countries. So, while the damaged human capital for unskilled labor can be ―restored‖,

the restoration process may accelerate the poverty problem (because idle labor need to

stick with low wage jobs, intensify rural-urban migration, etc.). Albala-Bertrand

(1993a) pointed this out that ―long-range effects of disaster situation do not primarily

depend on disaster loss but on its interference with on going social dynamics‖ (p. 204).

This is also the point that studies on the long-run impact of disaster are highly desired.

4.3. Negative Externalities

18

Natural disasters are different events from traffic accidents or even pandemics,

in that they can lead to a large number of simultaneous tragic deaths and injuries and

leave the affected people feeling hopeless against the hazard per se7. This can in turn

create disaster-specific negative externalities to the affected economy and/or

surrounding regions. Because of the extent of catastrophe, even survivors in the

damaged area are often affected psychologically, such as depression and/or

post-traumatic stress disorders (PTSD), and these conditions may lead to long term

adverse consequences, for example decreased productivity. In addition to the personal

impact from threat (hazard), Norris (2005) found through a review of the previous

studies that community destruction was more strongly related to decrease in regaining

positive motivation, leading the affected to feel less positive about the community, less

enthusiastic, less energetic, and less able to enjoy life, and concluded that natural

disasters in developing countries often produce severe effects on the public‘s mental

health. This may lead to losses in productivity, and hence an indication of a

disaster-specific negative externality. On the other hand, based on a survey of Sri

Lanka micro-enterprise owners affected by the 2004 Indian Ocean Earthquake and

Tsunami, de Mel et al., (2008) found that ―mental health recovery is largely a function

of time since the disaster, and is not strongly related to individual-level economic

recovery‖ (p. 514). However, the economic impact of negative externalities based on

psychological effects is largely unknown, in terms of how and to what extent it affects

economic activities, and needs to be further investigated.

After the 1995 Kobe Earthquake, Miyao (1995) suggested that the increase in

lay-offs and unemployment, the decreased value of wealth, and/or the decrease of

consumption by depression as the consequences of disasters, may have significant

effects throughout the economy. The first two effects can be incorporated and

evaluated with flow analysis models and stock analysis models, respectively, as

discussed above. However, the third effect, decreased consumption by depression,

which not only can take place in the damaged area but also can spread to the

surrounding areas, has been often overlooked. Okuyama et al. (1999) reported this

7 Again, while the occurrence of hazards cannot be prevented or even reduced, the extent of disaster can

be mitigated with appropriate countermeasures and preparedness. This is the main reason for

understanding the full scale of disaster impact.

19

effect in the case of the 1995 Kobe Earthquake, Japan, and Hewings and Okuyama

(2003) also reported the consumption slump after the terrorist attacks on September 11,

2001 in the United States. In these cases, consumers delayed purchases of major items

because of the depression and/or catastrophic atmosphere felt after the event, and this

turned out to be critical for the economic conditions during the recovery period.

Empirical studies on decreased consumption in a disaster situation in developing

countries was investigated by Auffret (2003), in which he proposed the framework to

analyze the effects of catastrophic events on household and aggregate welfare and tested

it with the Caribbean and Latin American countries for 1970-99. While the Caribbean

countries have suffered a high volatility of consumption over the years, his framework

suggested that the volatility of consumption comes from production shocks by

catastrophic events and the shocks are transformed into consumptions shocks, mostly

due to underdeveloped or ineffective risk management mechanisms in those countries.

Consumption decline or volatility after a disaster in the Auffret‘s case may be caused

through endogenous factors in an economy, rather than externalities; however, it may

create further depressed atmosphere (negative externality) where the disaster already

brought calamity.

In summarizing the above negative externalities discussed here, the effects can

be categorized to three-fold: personal losses and depression caused by the occurrence of

a hazard can be considered as direct psychological effect; the community-level

psychological impact as the supply (production) side effect, or indirect psychological

effect; and, the decreased consumption as the demand (consumption) side effect, or

secondary psychological effect. While some other negative externalities can be

identified, above issues need to be included in the evaluation of disaster impacts for a

more comprehensive understanding of the event‘s impact.

5. Conclusions

Recent frequent and severe disasters around the world have increased the interests and

urgency of researchers and policymakers to investigate and understand the extent and

significance of economic impacts. Disaster, as a phenomenon, is quite different from

20

any other economic events, like the recent global financial crisis, in many ways: sudden

occurrence of the event; loss of lives; negative externalities; non-uniform distribution of

damages over space; and a dynamic trend change from large negative shocks to a

positive demand injection for recovery and reconstruction. Therefore, the impact

estimate of disasters is very delicate methodologically, since standard economic models

assume gradual, incremental, and uniform changes. Yet, as reviewed above, there are

methodologies in use for disaster impact analysis, such as IO, SAM, CGE, and

econometric models, with their attendant strengths and weaknesses, for example

overestimation of impact in IO and SAM and underestimation in CGE. These

methodologies have advanced significantly through modification and extension of their

frameworks and/or integration with other models, to address specific features.

However, challenges, such as the assessment of long run impact and incorporation of

negative externalities, remain, especially in the developing country context.

The above discussions are, however, from the viewpoint of researchers, i.e. the

supply side of methodologies. The demand, or user, side of methodologies may often

have slightly different perspectives (see discussions in Okuyama, 2007), and it is

important to see how the two perspectives can reconcile the differences. Analysis of

disaster impact can be utilized by many users as follows:

a) emergency management officials (ex-post response);

b) policy makers for preparedness and mitigation policies (ex-ante response);

c) development planners; and

d) a wide range of researchers who are/will be engaging the disaster impact

analysis.

Flow analysis models can be utilized by a) emergency management and b) policy

makers to grasp the size and extent of disaster effects within a short period of time. In

this case, the users do not require the detail or sophistication of methodology per se;

rather, they demand the credibility (not accuracy) of the impact estimate. At the same

time, a) emergency management would also like to find the severity and extent of

negative externalities from psychological impact so that appropriate response, such as

dispatching psychiatrists for providing medical treatment, can be made in the damaged

area.

Long run impact analysis is necessary for the decision making of c)

21

development planners. Especially in developing countries, physical recovery from a

disaster may involve development implications (Albala-Bertrand, 1993a). As

indicated in the previous section, stock analysis models are desperately craved for both

empirical analysis and theoretical progress. At the same time, the connection between

disaster-induced development and other development issues, such as poverty and

structural inequality, must be also carefully investigated. Meanwhile, in order to

advance our understanding of disasters and their impacts, it is essential that d) a wide

range of researchers cooperate together. Issues discussed in the previous section, such

as long run impact and negative externalities, especially in the developing country

context, should receive more attention.

All the above discussions here, in turn, lead toward the necessity of reliable data

collection of disasters, and it appears now that having standardized definitions and data

of damages and losses for all the disasters remains a dream. West and Lenze (1994)

suggested that the more sophisticated impact models become, the more precise

numerical data will be required, while imperfect measurements of the damages and

losses of a disaster are often the case. In international settings, definitions of damages

and losses are often different across countries, let alone the definition of economic

impacts. As economic structures within a country and between countries have become

much more complex and intertwined, the methodologies for both flow and stock

analyses have also been more sophisticated. Whereas the demand for details are

different among the users discussed above, disaster data have to be standardized,

reliable, and consistent to some extent for all disasters in order to make cross-disaster

analysis. While collecting such data retrospectively for the past cases is hardly

affordable and unrealistic, it is never too late to set and implement some international

standards for future events, such as the method devised by the Economic Commission

for Latin American and the Caribbean (ECLAC, 2003). On this front, it may also be

useful to link economic models for disaster with other established indicators of

development, such as the Economic Vulnerability & Resilience Monitor (Econ-VR

Monitor) of the United Nation‘s International Strategy for Disaster Reduction

(UN-ISDR), so that the data collection and analysis can be streamlined and linked

together, and that the models and the resulting analyses will be much more useful to the

disaster community, researchers, and practitioners a like.

22

References

Albala-Bertrand, J.M. (1993a) The Political Economy of Large Natural Disasters: With

Special Reference to Developing Countries (Oxford, UK: Clarendon Press).

Albala-Bertrand, J.M. (1993b) Natural disaster situations and growth: a macroeconomic

model for sudden disaster impacts, World Development, 21 (9), pp.1417-1434.

Anbarci, N., Escalerasb, M., and Register, C.A. (2005) Earthquake fatalities: the

interaction of nature and political economy, Journal of Public Economics, 89, pp.

1907–1933.

Ariyabandu, M.M. (2001) Bringing together disaster and development - concepts and

practice, some experience from South Asia, Paper presented at the 5th European

Sociological Association Conference, in Helsinki, August 28th-September 1st,

2001.

Auffret, P. (2003) High consumption volatility: the impact of natural disasters? Policy

Research Working Paper 2962, Washington DC: World Bank.

Benson, C. and Clay, E. (1998) The impact of drought on sub-Saharan African

economies, Technical paper 401, Washington DC: World Bank.

Boisvert, R. (1992) Indirect losses from a catastrophic earthquake and local, regional,

and national interest, in: Indirect Economic Consequences of a Catastrophic

Earthquake, pp. 207-265 (Washington, DC: National Earthquake Hazards

Reduction Program, Federal Emergency Management Agency).

Brookshire, D. and Mckee, M. (1992) Other indirect costs and losses from earthquakes:

issues and estimation, in: Indirect Economic Consequences of a Catastrophic

Earthquake, pp. 267-325 (Washington, DC: National Earthquake Hazards

Reduction Program, Federal Emergency Management Agency).

Bruneau, M., Chang, S.E., Eguchi, R., Lee, G., O‘Rourke, T., Reinhorn, A., Shinozuka,

M., Tierney, K., Wallace, W. and von Winterfelt, D. (2003) A framework to

quantitatively assess and enhance seismic resilience of communities, Earthquake

Spectra, 19, pp. 733-752.

Cardenasa, V., Hochrainerb, S., Mechlerb, R., Pflugb, G., and Linnerooth-Bayerb, J.

(2007) Sovereign financial disaster risk management: The case of Mexico,

Environmental Hazards, 7, pp. 40-53.

Cho, S., Gordon, P., Moore II, J.E., Richardson, H.W., Shinozuka, M. and Chang, S.E.

(2001) Integrating transportation network and regional economic models to estimate

the costs of a large urban earthquake, Journal of Regional Science, 41, pp. 39-65.

Cochrane, H.C. (1974) Predicting the economic impacts of earthquakes, in: H.C.

Cochrane, J.E. Haas and R.W. Kates (Eds) Social Science Perspectives on the

23

Coming San Francisco Earthquakes—Economic Impact, Prediction, and

Reconstruction, Natural Hazard Working Paper No.25 (Boulder, CO: University of

Colorado, Institute of Behavioral Sciences).

Cochrane, H.C. (1997) Forecasting the economic impact of a Midwest earthquake, in:

B.G. Jones (Ed.) Economic Consequences of Earthquakes: Preparing for the

Unexpected, pp. 223-247 (Buffalo, NY: National Center for Earthquake

Engineering Research).

Cochrane, H.C., Chang, S.E. and Rose, A. (1997) Indirect economic losses, in:

Earthquake Loss Estimation Methodology: HAZUS Technical Manual, Volume 3

(Washington DC: National Institute of Building Sciences).

Cole, S. (1988) The delayed impacts of plant closures in a reformulated Leontief model,

Papers of the Regional Science Association, 65, pp. 135-149.

Cole, S. (1989) Expenditure lags in impacts analysis, Regional Studies, 23, pp. 105-116.

Cole, S. (1995) Lifeline and livelihood: a social accounting matrix approach to calamity

preparedness, Journal of Contingencies and Crisis Management, 3, pp. 228-40.

Cole, S. (1998) Decision support for calamity preparedness: socioeconomic and

interregional impacts, in: M. Shinozuka, A. Rose and R.T. Eguchi (Eds)

Engineering and Socioeconomic Impacts of Earthquakes, pp. 125-153 (Buffalo,

NY: Multidisciplinary Center for Earthquake Engineering Research).

Cole, S. (2004) Geohazards in social systems: an insurance matrix approach, in: Y.

Okuyama and S.E. Chang (Eds) Modeling Spatial and Economic Impacts of

Disasters, pp. 103-118 (New York: Springer).

Cuaresma, J.C., Hlouskova, J., and Obersteiner, M. (2008) Natural disasters as creative

destruction?: evidence from developing countries, Economic Inquiry, 46 (2), pp.

214-226.

Dacy , D.C. and Kunreuther, H. (1969) The Economics of Natural Disasters (New York,

The Free Press).

Davis, D.R. and Weinstein, D.E. (2002) Bones, bombs, and breakpoints: the geography

of economic activity, American Economic Review, 92 (5), pp. 1269-1289.

Davis, D.R. and Weinstein, D.E. (2004) A search for multiple equilibria in urban

industrial structure, Working Paper 10252, National Bureau of Economic Research.

Davis, H.C. and Salkin, E.L. (1984) Alternative approaches to the estimation of

economic impacts resulting from supply constraints, Annals of Regional Science, 18,

pp. 25-34.

De Mel, S., McKenzie, D., and Woodruff, C. (2008) Mental health recovery and

economic recovery after the tsunami: high-frequency longitudinal evidence from Sri

Lankan small business owners, Social Science & Medicine, 66 (3), pp. 582–595.

Dietzenbacher, E. (1997) In vindication of the Ghosh model: a reinterpretation as a

price model, Journal of Regional Science, 37, pp. 629-651

Donaghy, K.P., Balta-Ozkan, N., Hewings, G.J.D. (2007) Modeling unexpected events

in temporally disaggregated econometric input-output models of regional

24

economies, Economic Systems Research, 19 (2), pp. 125-145.

Economic Commission for Latin American and the Caribbean (ECLAC) (2003)

Handbook for Estimating the Socio-economic and Environmental Effects of

Disasters.

Ellson, R.W., Milliman, J.W. and Roberts, R.B. (1984) Measuring the regional

economic effects of earthquakes and earthquake predictions, Journal of Regional

Science, 24, pp. 559-579.

Freeman, P.K., Martin, L.A., Mechler, R., Warner, K., and Hausmann, P. (2002)

Catastrophes and Development: Integrating Natural Catastrophes into Development

Planning, Working Paper Series No. 4, World Bank.

Gordon, P. and Richardson, H.W. (1996) The business interruption effects of the

Northridge earthquake (Lusk Center Research Institute, University of Southern

California, Los Angeles, CA).

Gordon, P., Richardson, H.W. and Davis, B. (1998) Transport-related impacts of the

Northridge earthquake, Journal of Transportation and Statistics, 1, pp. 22-36.

Hallegatte, S. and Ghil, M. (2008) Natural disasters impacting a macroeconomic model

with endogenous dynamics, Ecological Economics, 68, pp. 582-592.

Hewings, G.J.D. and Okuyama, Y. (2003) Economic assessment of unexpected events,

in S.L. Cutter, D.B. Richardson, and T.J. Wilbanks (Eds) The Geographical

Dimensions of Terrorism, Routledge, pp. 153-160.

Hoke, Z. (2005) In natural disasters the poor are hardest hit, Voice of America News,

http://voanews.com/english/archive/2005-01/2005-01-13-voa12.cfm, accessed on

10/16/08.

Ikefuji, M. and Horii, R. (2006) Natural disasters in a two-sector of endogenous model,

Discussion Paper 06-13, Discussion Papers in Economics and Business, Osaka

University.

Kahn, M.E. (2005) The death toll from natural disasters: the role of income, geography,

and institutions, The Review of Economics and Statistics, 87 (2), pp. 271–284.

Kajitani, Y., Tatano, H., Yamano, N., and Shumuta, Y. (2005a) Estimation of resiliency

factor of industrial sectors under multiple lifeline disruptions, Natural Disaster

Science, 23 (4), pp. 553-564 (in Japanese).

Kajitani, Y., Tatano, H., Yamano, N., and Shumuta, Y. (2005b) Estimation of resiliency

factor of non-industrial sectors under multiple lifeline disruptions, Natural Disaster

Science, 24 (3), pp. 247-255 (in Japanese).

Kawashima, K., Sugita, H. and Kanoh, T. (1991) Estimation of earthquake induced

economic damage, Report of Public Works Research Institute, 186, pp. 1-57. (In

Japanese.)

Kellenberg, D.K. and Mobarak, A.M. (2008) Does rising income increase or decrease

damage risk from natural disasters? Journal of Urban Economics, 63, pp. 788-802.

Kunreuther, H. and Rose, A. (2004b) Introduction, in: H. Kunreuther and A. Rose (Eds)

The Economics of Natural Hazards, Volume I, pp. xiii-xxxv (Northampton, MA:

25

Edward Elgar).

Kunreuther, H. and Rose, A. (Eds) (2004a) The Economics of Natural Hazards, Volume

I & II (Northampton, MA: Edward Elgar).

Lester, R. (2008) Climate change, development and the insurance sector‖, Policy

Research Working Paper, World Bank.

Miguel, E. and Roland, G. (2005) The long run impact of bombing Vietnam, mimeo,

University of California, Berkeley.

Miyao, T. (1995) Reconstruction boom after the Great Hanshin Earthquake won‘t arise,

Economist (Japanese edition), 3/7/95 issue, pp/ 26-30 (in Japanese).

Norris, F.H. (2005) Psychosocial consequences of natural disasters in developing

countries: what does past research tell us about the potential effects of the 2004

tsunami? National Center for PTSD, Dartmouth College,

http://www.ncptsd.va.gov/ncmain/ncdocs/fact_shts/fs_tsunami_research.html?print

able-template=factsheet, accessed on 08/04/2008.

Okonski, K. (2004) Economic growth saves lives, Wall Street Journal Asia, December

29, 2004.

Okuyama, Y. (2003) Economics of natural disasters: a critical review, Research Paper

2003-12, Regional Research Institute, West Virginia University.

Okuyama, Y. (2006) Sequential interindustry model and inventory function: sensitivity

analysis, Paper presented at the 2006 Intermediate Input-Output Meetings on

Sustainability, Trade & Productivity, Sendai, Japan.

Okuyama, Y. (2007) Economic modeling for disaster impact analysis: past, present, and

future, Economic Systems Research, 19 (2), pp. 115-124.

Okuyama, Y. and Chang, S.E. (2004b) Introduction, in: Y. Okuyama and S.E. Chang

(Eds) Modeling Spatial and Economic Impacts of Disasters, pp. 1-10 (New York:

Springer).

Okuyama, Y. and Chang, S.E. (Eds) (2004a) Modeling Spatial and Economic Impacts of

Disasters (New York: Springer).

Okuyama, Y., Hewings, G.J.D. and Sonis, M. (1999) Economic impacts of an

unscheduled, disruptive event: a Miyazawa multiplier analysis, in G.J.D. Hewings,

M. Sonis, M. Madden and Y. Kimura (Eds) Understanding and Interpreting

Economic Structure, pp. 113-144 (Berlin: Springer).

Okuyama, Y., Hewings, G.J.D. and Sonis, M. (2004) Measuring economic impacts of

disasters: interregional input-output analysis using sequential interindustry model,

in: Y. Okuyama and S.E. Chang (Eds) Modeling Spatial and Economic Impacts of

Disasters, pp. 77-101 (New York: Springer).

Oosterhaven, J. (1988) On the plausibility of the supply-driven input-output model,

Journal of Regional Science, 28, pp. 203-217.

Oosterhaven, J. (1989) The supply-driven input-output model: a new interpretation but

still implausible, Journal of Regional Science, 29, pp. 459-465.

26

Pelling, M., Özerdem, A., and Barakat, S. (2002) The macro-economic impact of

disasters, Progress in Development Studies, 2 (4), pp. 283-305.

Rasmussen, T.N. (2004) Macroeconomic implications of natural disasters in the

Caribbean, IMF Working Paper WP/04/224, International Monetary Fund.

Romanoff, E. and Levine, S.H. (1986) Capacity limitations, inventory, and time-phased

production in the sequential interindustry model, Papers of the Regional Science

Association, 59, pp. 73-91.

Rose, A. (1981) Utility lifelines and economic activity in the context of earthquakes, in

J. Isenberg (Ed.) Social and Economic Impacts of Earthquake on Utility Lifelines,

pp. 107-120 (New York: American Society of Civil Engineers).

Rose, A. (2004) Economic principles, issues, and research priorities in hazard loss

estimation, in: Y. Okuyama and S.E. Chang (Eds) Modeling Spatial and Economic

Impacts of Disasters, pp. 13-36 (New York: Springer).

Rose, A. and Benavides, J. (1998) Regional economic impacts, in: M. Shinozuka, A.

Rose and R.T. Eguchi (Eds) Engineering and Socioeconomic Impacts of

Earthquakes, pp. 95-123 (Buffalo, NY: Multidisciplinary Center for Earthquake

Engineering Research).

Rose, A. and Guha, G.S. (2004) Computable general equilibrium modeling of electric

utility lifeline losses from earthquakes, in: Y. Okuyama and S.E. Chang (Eds)

Modeling Spatial and Economic Impacts of Disasters, pp. 119-141 (New York:

Springer).

Rose, A. and Liao, S.Y. (2005) Modeling regional economic resilience to disasters: a

computable general equilibrium analysis of water service disruptions, Journal of

Regional Science, 45, pp. 75-112.

Rose, A., Benavides, J., Chang, S.E., Szczesniak, P. and Lim, D. (1997) The regional

economic impact of an earthquake: direct and indirect effects of electricity lifeline

disruptions, Journal of Regional Science, 37, pp. 437-458.

Shah, S. (1999) Coping with natural disasters: the 1998 floods in Bangladesh, a paper

presented in 1999 World Bank Summer Research Workshop on Poverty and

Development, July 6-8, 1999, Washington, D.C.

Shinozuka, M. and Chang, S.E. (2004) Evaluating the disaster resilience of power

networks and grids, in: Y. Okuyama and S.E. Chang (Eds) Modeling Spatial and

Economic Impacts of Disasters, pp. 289-310 (New York: Springer).

Shumuta, Y. (2004) Benefit cost analysis for renewal planning of existing electric power

equipment, in: Y. Okuyama and S.E. Chang (Eds) Modeling Spatial and Economic

Impacts of Disasters, pp. 257-287 (New York: Springer).

Skidmore, M. and Toya, H. (2002) Do natural disasters promote long-run growth?

Economic Inquiry, 40 (4), pp. 664-687.

Sohn, J., Hewings, G.J.D., Kim, T.J., Lee, J.S. and Jang, S.G. (2004) Analysis of

economic impacts of earthquake on transportation network, in: Y. Okuyama and S.E.

Chang (Eds) Modeling Spatial and Economic Impacts of Disasters, pp. 233-256

(New York: Springer).

27

Sorkin, A.L. (1982) Economic Aspects of Natural Hazards (Lexington, MA: Lexington

Books).

Steinback, S.R. (2004) Using ready-made regional input-output models to estimate

backward-linkage effects of exogenous output shocks, Review of Regional Studies,

34, pp. 57-71.

Tierney, K. (1997) Impacts of recent disasters on business: the 1993 Midwest floods and

the 1994 Northridge earthquake, in: B.G. Jones (Ed) Economic Consequences of

Earthquakes: Preparing for the Unexpected, pp. 189-221 (Buffalo, NY: National

Center for Earthquake Engineering Research).

Tol, R. and Leek, F. (1999) Economic analysis of natural disasters, in T. Downing, A.

Olsthoorn, and R. Tol (Eds) Climate Change and Risk, London: Routledge,

pp.308-327.

Toya, H. and Skidmore, M. (2005) Economic development and the impacts of natural

disasters, Working Paper 05-04, Department of Economics, University of Wisconsin

– Whitewater.

Tsuchiya, S., Tatano, H., and Okada, N. (2007) Economic loss assessment due to

railroad and highway disruptions, Economic Systems Research, 19 (2), pp. 147-162.

Van der Veen, A. and Logtmeijer, C.. (2003) How vulnerable are we for flooding? A

GIS approach, in: A. van der Veen, A.L.V. Arellano and J.P. Nordvik (Eds) In

Search of a Common Methodology on Damage Estimation (EUR 20997 EN), pp.

181-193 (Ispra, Italy: European Commission, Joint Research Center).

West, C.T. and Lenze, D.G. (1994) Modeling the regional impact of natural disaster and

recovery: a general framework and an application to hurricane Andrew,

International Regional Science Review, 17, pp. 121-150.

Wilson, R. (1982) Earthquake vulnerability analysis for economic impact assessment

(Information Resources Management Office, Federal Emergency Management

Agency, Washington, DC).

Yamano, N., Kajitani, Y., and Shumuta, Y. (2007) Modeling the regional economic loss

of natural disasters: the search for economic hotspots, Economic Systems Research,

19 (2), pp. 163-181.