Climate change futures: health, ecological and economic dimensions

Climate signals in rising costs from “natural” disasters
are evident in many aspects of the data. Insurers
observe a notable increase in losses during periods of
elevated temperatures 4 and lightning strikes (predicted
to rise with warming) (Price and Rind 1993,
1994a,b; Reeve and Toumi 1999) (see figure 3.6).
Other factors, not captured by an overall examination
of losses, point to an inherent bias toward underreporting
the economic impact of climate extremes. The total
losses are underestimates, since record-keeping systematically
ignores relatively small events. 5 For example,
smaller, often uncounted losses include those from soil
subsidence or permafrost melt.
In the summer of 2005, for example, there were
numerous areas of drought, flashfloods and wildfires
that were barely accounted for.
United States alone are estimated to result in a cost of
US $80 billion per year (LaCommare and Eto 2004),
and weather-related events account for 60% of the customers
affected by grid disturbances in the bulk power
markets (see figure 1.8). Lightning strikes collectively
result in billions of dollars of losses, as do damages to
human infrastructure from soil subsidence (Nelson et al.
2001). Yet, such small-scale events are rarely if ever
included in US insurance statistics due to the minimum
event cost threshold of US $5 million up to 1996, and
US $25 million thereafter. The published insurance figures
reflect property losses and largely exclude the loss
of life, and health costs (which are diffuse and rarely
tabulated), business interruptions, restrictions on trade,
travel and tourism, and potential market instability
resulting from the health and ecological consequences
of warming temperatures and severe weather.
24 | THE CLIMATE CONTEXT TODAY
While attention naturally focuses on headline-grabbing
catastrophes, the majority (60%) of economic losses
comes from smaller events (Mills 2005). In one recent
example, a month of extremely cold weather in the
northeastern US in 2004 resulted in US $0.725 billion
in insured losses (American Re 2005). The average
annual insured loss from winter storms and thunderstorms
is about US $6 billion in the US, comparable
to the loss from a significant hurricane (American
Re 2005).
In addition, while the absolute magnitude of losses has
been rising, the variability of losses has also been
increasing in tandem with more variability in weather.
LOSSES ARE SYSTEMATICALLY
UNDERESTIMATED
The magnitude of losses presented in published data
systematically underestimates the actual costs. For
example:
Statistical bodies commonly create definitions that
exclude from tabulation those events falling below a
given threshold. For example, power outages in the
4. Data furnished by Richard Jones, Hartford Steam Boiler Insurance and
Inspection Service (see figure 3.6)
5. For example, the insurance industry's Property Claim Services tabulated
only those losses of US $5 million or more up until 1996 and those of US
$25 million or more thereafter. As a result, no winter storms were included in
the statistics for the 26-year period of 1949-1974, and few were thereafter
(Kunkel et al. 1999). Although large in aggregate, highly diffuse losses due to
structural damages from land subsidence would also rarely be captured in
these statistics. Similarly, weather-related vehicular losses are typically not captured
in the statistics.
The published data are not necessarily directly comparable
with past data. For instance, in recent years,
there has been a trend toward both increasing
deductibles and decreasing limits, resulting in lower
insurance payouts than had the rules been unchanged.
Following Hurricane Andrew, insurers instituted special
“wind” deductibles, in addition to the standard property
deductible now in use in 18 US states plus
Washington, DC (Green 2005a,b). Moreover, hurricane
deductibles have moved toward a percentage of
the total loss rather than the traditionally fixed formulation.
The effect of such changes is substantial: for
example, in Florida,15 to 20% of the losses from the
2004 hurricanes were borne by consumers (Musulin
and Rollins 2005).
These data also, of course, exclude the costs for disaster
preparedness or adaptation to the rise in extreme
weather events (flood preparedness, changes in construction
practices and codes, improved fire suppression
technology, cloud seeding, lightning protection,
etc.). Insurance industry representatives reported that
improved building codes helped reduce the losses of
hurricanes in 2004 (Khanduri 2004).
Other countervailing factors also mask part of the
actual upward pressure on costs. Improved building
codes, early warning systems, river channelization,
cloud seeding and fire suppression all offset losses that
might otherwise have occurred. Financial factors, such
as insurer withdrawal from risky areas, higher
deductibles and lower limits, also have a dampening
effect on losses. All this means that the data do not
necessarily reflect the increased costs to society of a
changing climate.