Geoscience Reference
In-Depth Information
consequences of policy interventions. While regional CGE modeling has a three
decade history of providing input to evidence-based policy making, research
exploiting the exogenous structural, behavioral and policy detail in these models
to carry shocks describing the direct effects of catastrophic events is more recent.
1
Early applications examined natural disasters. Rose and Liao (
2005
) examined the
effects of water utility disruption following an earthquake, noting the role of
pre-event mitigation and post-event resilience in influencing potential impacts.
More recently, the CGE method has been turned to terrorism threats. Rose
et al. (
2009
) examined the consequences of the 2001 World Trade Center attacks,
investigating BI costs, and behavioral impacts via reduced air travel. The impor-
tance of behavioral responses within the overall economic consequences of a
terrorist event was examined more generally in Giesecke et al. (
2012
). Investigating
the consequences of a radiological dispersion device (RDD) attack in the financial
district of downtown Los Angeles (LA), they compare economic costs arising from
behavioral responses with the direct resource costs arising from casualties, property
damage and BI.
In this paper we use a dynamic regional CGE model to perform a consequence
analysis of a terrorist chlorine attack with a focus on behavioral impacts and
dynamic outcomes. We perform a decomposition of the event's various loss
components to explain and compare behavioral impacts with more standard
resource loss effects over time. Given the potential for multiple economic loss
channels, risk managers would be well advised to distinguish the many types of
consequences of a terrorist attack or natural disaster. Unfortunately, most conse-
quence analyses have not offered such decomposition analyses of the broad range
of components. Furthermore, whereas previous studies have used comparative
static models, our model is dynamic. A dynamic model offers a number of benefits
relative to its comparative static counterpart: first, by providing a more plausible
time-path for key regional stock variables (like population and capital) by allowing
for gradual adjustment in regional wages, migration, and investment; and second,
by facilitating a better matching of time-specific inputs to the CGE model with the
projected time-path of peak and decay in behavioral responses to a particular hazard
or disaster.
To provide a comparison with the scenario in Giesecke et al. (
2012
), we choose
the same downtown LA area as the attack site: the heart of the financial district. The
attack involves detonation of a chlorine storage tank, leading to the formation of a
large chlorine plume. Consistent with the approach outlined in Giesecke et al., we
divide the event's direct effects into two broad sets of inputs to the CGE model:
(1) reduction in effective resource supply (the resource loss effect) and (2) shifts in
the perceptions of economic agents (the behavioral effect). The resource loss effect
describes the event's physical destructiveness, subsuming such direct impacts as
deaths and injuries, BI during evacuation, clean-up and repair, and medical
1
Recent reviews of regional CGE model applications to the analysis of public policy and regional
development are provided by Partridge and Rickman (
2010
) and Giesecke and Madden (
2013
).
