Climate Change
Financial Risks to Federal and Private Insurers in Coming Decades Are Potentially Significant Gao ID: GAO-07-285 March 16, 2007Weather-related events have cost the nation billions of dollars in damages over the past decade. Many of these losses are borne by private insurers and by two federal insurance programs--the National Flood Insurance Program (NFIP), which insures properties against flooding, and the Federal Crop Insurance Corporation (FCIC), which insures crops against drought or other weather disasters. GAO was asked to (1) describe how climate change may affect future weather-related losses, (2) determine past insured weather-related losses, and (3) determine what major private insurers and federal insurers are doing to prepare for potential increases in such losses. In response, among other things, GAO reviewed key scientific assessments; analyzed insured loss data; and contacted private insurers, NFIP, and FCIC.
Key scientific assessments report that the effects of climate change on weather-related events and, subsequently, insured and uninsured losses, could be significant. The global average surface temperature has increased by 0.74 degrees Celsius over the past 100 years and climate models predict additional, perhaps accelerating, increases in temperature. The key assessments GAO reviewed generally found that rising temperatures are expected to increase the frequency and severity of damaging weather-related events, such as flooding or drought, although the timing and magnitude are as yet undetermined. Additional research on the effect of increasing temperatures on weather events is expected in the near future, including a highly anticipated assessment of the state of climate science this year. Taken together, private and federal insurers paid more than $320 billion in claims on weather-related losses from 1980 to 2005. Claims varied significantly from year to year--largely due to the effects of catastrophic weather events such as hurricanes and droughts--but have generally increased during this period. The growth in population in hazard-prone areas and resulting real estate development have generally increased liabilities for insurers, and have helped to explain the increase in losses. Due to these and other factors, federal insurers' exposure has grown substantially. Since 1980, NFIP's exposure quadrupled, nearing $1 trillion in 2005, and program expansion increased FCIC's exposure 26-fold to $44 billion. Major private and federal insurers are both exposed to the effects of climate change over coming decades, but are responding differently. Many large private insurers are incorporating climate change into their annual risk management practices, and some are addressing it strategically by assessing its potential long-term industry-wide impacts. The two major federal insurance programs, however, have done little to develop comparable information. GAO acknowledges that the federal insurance programs are not profit-oriented, like private insurers. Nonetheless, a strategic analysis of the potential implications of climate change for the major federal insurance programs would help the Congress manage an emerging high-risk area with significant implications for the nation's growing fiscal imbalance.
RecommendationsOur recommendations from this work are listed below with a Contact for more information. Status will change from "In process" to "Open," "Closed - implemented," or "Closed - not implemented" based on our follow up work.
Director: Team: Phone:GAO-07-285, Climate Change: Financial Risks to Federal and Private Insurers in Coming Decades Are Potentially Significant
This is the accessible text file for GAO report number GAO-07-285
entitled 'Climate Change: Financial Risks to Federal and Private
Insurers in Coming Decades Are Potentially Significant' which was
released on April 19, 2007.
This text file was formatted by the U.S. Government Accountability
Office (GAO) to be accessible to users with visual impairments, as part
of a longer term project to improve GAO products' accessibility. Every
attempt has been made to maintain the structural and data integrity of
the original printed product. Accessibility features, such as text
descriptions of tables, consecutively numbered footnotes placed at the
end of the file, and the text of agency comment letters, are provided
but may not exactly duplicate the presentation or format of the printed
version. The portable document format (PDF) file is an exact electronic
replica of the printed version. We welcome your feedback. Please E-mail
your comments regarding the contents or accessibility features of this
document to Webmaster@gao.gov.
This is a work of the U.S. government and is not subject to copyright
protection in the United States. It may be reproduced and distributed
in its entirety without further permission from GAO. Because this work
may contain copyrighted images or other material, permission from the
copyright holder may be necessary if you wish to reproduce this
material separately.
Report to the Committee on Homeland Security and Governmental Affairs,
U.S. Senate:
United States Government Accountability Office:
GAO:
March 2007:
Climate Change:
Financial Risks to Federal and Private Insurers in Coming Decades Are
Potentially Significant:
GAO-07-285:
GAO Highlights:
Highlights of GAO-07-285, a report to the Committee on Homeland
Security and Governmental Affairs, U.S. Senate
Why GAO Did This Study:
Weather-related events have cost the nation billions of dollars in
damages over the past decade. Many of these losses are borne by private
insurers and by two federal insurance programs”the National Flood
Insurance Program (NFIP), which insures properties against flooding,
and the Federal Crop Insurance Corporation (FCIC), which insures crops
against drought or other weather disasters.
GAO was asked to (1) describe how climate change may affect future
weather-related losses, (2) determine past insured weather-related
losses, and (3) determine what major private insurers and federal
insurers are doing to prepare for potential increases in such losses.
In response, among other things, GAO reviewed key scientific
assessments; analyzed insured loss data; and contacted private
insurers, NFIP, and FCIC.
What GAO Found:
Key scientific assessments report that the effects of climate change on
weather-related events and, subsequently, insured and uninsured losses,
could be significant. The global average surface temperature has
increased by 0.74 degrees Celsius over the past 100 years and climate
models predict additional, perhaps accelerating, increases in
temperature. The key assessments GAO reviewed generally found that
rising temperatures are expected to increase the frequency and severity
of damaging weather-related events, such as flooding or drought,
although the timing and magnitude are as yet undetermined. Additional
research on the effect of increasing temperatures on weather events is
expected in the near future, including a highly anticipated assessment
of the state of climate science this year.
Taken together, private and federal insurers paid more than $320
billion in claims on weather-related losses from 1980 to 2005. Claims
varied significantly from year to year”largely due to the effects of
catastrophic weather events such as hurricanes and droughts”but have
generally increased during this period. The growth in population in
hazard-prone areas and resulting real estate development have generally
increased liabilities for insurers, and have helped to explain the
increase in losses. Due to these and other factors, federal insurers‘
exposure has grown substantially. Since 1980, NFIP‘s exposure
quadrupled, nearing $1 trillion in 2005, and program expansion
increased FCIC‘s exposure 26-fold to $44 billion.
Major private and federal insurers are both exposed to the effects of
climate change over coming decades, but are responding differently.
Many large private insurers are incorporating climate change into their
annual risk management practices, and some are addressing it
strategically by assessing its potential long-term industry-wide
impacts. The two major federal insurance programs, however, have done
little to develop comparable information. GAO acknowledges that the
federal insurance programs are not profit-oriented, like private
insurers. Nonetheless, a strategic analysis of the potential
implications of climate change for the major federal insurance programs
would help the Congress manage an emerging high-risk area with
significant implications for the nation‘s growing fiscal imbalance.
Figure: Growth in Exposure of Federal Insurance Programs ($2005):
[See PDF for Image]
Source: GAO.
[End of figure]
What GAO Recommends:
GAO is recommending that the Secretaries of Agriculture and Homeland
Security analyze the potential long-term fiscal implications of climate
change for the FCIC and the NFIP, respectively, and report their
findings to the Congress. In commenting on a draft of this report, the
two agencies agreed with the recommendation. The Departments of
Agriculture and Commerce made comments and suggestions on the
presentation of several findings. The Department of Energy elected not
to comment.
[Hyperlink, http://www.gao.gov/cgi-bin/getrpt?GAO-07-285].
To view the full product, including the scope and methodology, click on
the link above. For more information, contact John Stephenson at (202)
512-3841 or stephensonj@gao.gov.
[End of section]
Contents:
Letter:
Results in Brief:
Background:
Climate Change May Increase Losses by Altering the Frequency or
Severity of Weather-Related Events:
Insured Weather-Related Losses Have Been Sizeable, and Federal
Insurers' Exposure Has Grown Significantly:
Major Private and Public Insurers Differ in How They Manage
Catastrophic Risks Associated with Climate Change:
Conclusions:
Recommendation for Executive Action:
Agency Comments and Our Evaluation:
Appendix I: Scope and Methodology:
Scientific Literature:
Insured Loss Data:
Interviews with Major Insurers:
Appendix II: National Flood Insurance Program:
How the Program Works:
Risk Assessment Practices:
Program Funding:
Appendix III: Federal Crop Insurance Corporation:
How the Program Works:
Risk Assessment Practices:
Program Funding:
Appendix IV: Consensus Statement among Participants at 2006 Munich Re
Workshop:
Appendix V: Comments from the U.S. Department of Agriculture:
GAO Comments:
Appendix VI: Comments from the Department of Commerce:
GAO Comments:
Appendix VII: GAO Contact and Staff Acknowledgments:
Related GAO Products:
Tables:
Table 1: Selected IPCC Estimates of Confidence in Projected Changes in
Weather-Related Events:
Table 2: Insured Losses Associated with Hurricanes:
Table 3: Key Policy-Oriented Scientific Assessments Reviewed by GAO:
Figures:
Figure 1: Time Line of Key Scientific Assessments:
Figure 2: July 1993 Flood Damage at Chesterfield Airport in St. Louis,
Missouri:
Figure 3: Economic Damages by Hurricane Category for U.S. Hurricanes
Making Landfall, 1900-2005:
Figure 4: Annual Weather-and Nonweather-Related Insured Losses:
Figure 5: Weather-Related Losses Paid by Private Insurers:
Figure 6: Weather-Related Losses Paid by NFIP:
Figure 7: Weather-Related Losses Paid by FCIC:
Figure 8: NFIP Policies and Total Coverage:
Figure 9: FCIC Total Coverage:
Figure 10: Modeling Potential Catastrophe Losses:
Abbreviations:
AAA: American Academy of Actuaries:
AMO: Atlantic Multidecadal Oscillation:
CCSP: Climate Change Science Program:
FAIR: Fair Access to Insurance Requirements:
FEMA: Federal Emergency Management Agency:
FCIC: Federal Crop Insurance Corporation:
HUD: Department of Housing and Urban Development:
IPCC: Intergovernmental Panel on Climate Change:
NAIC: National Association of Insurance Commissioners:
NAS: National Academy of Sciences:
NFIP: National Flood Insurance Program:
NHC: National Hurricane Center:
NOAA: National Oceanic and Atmospheric Administration:
PCS: Property Claim Services:
RMA: risk Management Agency:
SAP: synthesis and assessment product:
SFIP: standard flood insurance policy:
USDA: U.S. Department of Agriculture:
[End of section]
United States Government Accountability Office:
Washington, DC 20548:
March 16, 2007:
The Honorable Joseph I. Lieberman:
Chairman:
The Honorable Susan M. Collins:
Ranking Member:
Committee on Homeland Security and Governmental Affairs:
United States Senate:
As the 2004 and 2005 hurricane seasons demonstrated, weather-related
events can devastate affected communities and individuals, and are
costly to the insurance industry, government disaster assistance
programs, and other relief organizations. Apart from the record-setting
losses experienced in 2005, weather-related events over the past decade
have cost the country tens of billions of dollars each year.
The property and casualty segment of the insurance industry, spanning
both the private and public sector, bears a large portion of weather-
related losses.[Footnote 1] The private sector includes primary
insurers that insure individuals and businesses directly, and
reinsurers that provide insurance to the primary insurers. The public
sector includes federal programs--in particular, the National Flood
Insurance Program (NFIP), which insures properties at risk of damage
from flooding, and the Federal Crop Insurance Corporation (FCIC), which
insures crops that are vulnerable to drought, floods, or other natural
disasters. Many states also administer insurance pools that provide
coverage for losses caused by weather-related events.
The uncertain and potentially large losses associated with weather-
related events are among the biggest risks that property insurers face.
Virtually anything that is insured--property, crops and livestock,
business operations, or human life and health--is vulnerable to weather-
related events. To remain financially solvent, the insurance industry
must estimate and prepare for the potential impact of weather- related
events. As such, any unanticipated changes in the frequency or severity
of weather-related events can have financial consequences at the
company level and industry-wide.
The earth's climate and weather patterns are dynamic, varying on
seasonal, decadal, and longer time scales. The global average surface
temperature has increased by 0.74 degrees Celsius over the past 100
years and climate models predict additional, perhaps accelerating,
increases in temperature. While the temperature increases to date may
appear small, climate models project that additional changes in
temperature may alter social and economic activities in potentially
profound ways. Much research and policy debate has centered on the
extent to which human activities have contributed to the warming and
how much is due to natural variability. For the purposes of this
report, climate change refers to any change in the climate over time,
whether due to natural variability or as a result of human
activity.[Footnote 2] Regardless of the cause, some contend that
increasing temperatures--accompanied by changes in other aspects of the
climate--may have adverse financial consequences for property insurers,
which might slow the growth of the industry and shift more of the
burden to governments and individuals.
Concerned about the implications of climate change for weather-related
losses incurred by federal agencies and private insurers, you asked us
to (1) describe what is known about how climate change might affect
insured and uninsured losses, (2) determine insured losses incurred by
major federal agencies and private insurers and reinsurers resulting
from weather-related events, and (3) determine what major federal
agencies and private insurers and reinsurers are doing to prepare for
the potential risk of increased losses due to more frequent or more
severe weather-related events associated with climate change.
To describe how climate change might affect insured and uninsured
losses, we reviewed and summarized key scientific assessments by
reputable international and national research organizations, including
the Intergovernmental Panel on Climate Change Third Assessment Report,
National Academy of Sciences reports, and the multifederal agency
Climate Change Science Program. To determine insured losses
attributable to weather-related events, we analyzed data from 1980
through 2005 from the Department of Homeland Security's Federal
Emergency Management Agency (FEMA) for the NFIP; from the Department of
Agriculture's Risk Management Agency (RMA) for FCIC; and from the
Property Claims Service, a leading source of insurance data. We
analyzed changes in weather-related losses since 1980 and supplemented
this analysis with a review of existing literature and the views of
subject area experts on the key drivers of changes in losses.
To determine what key federal agencies and private insurers are doing
to assess and manage the potential for increased losses, we conducted
semistructured interviews with officials from the NFIP, RMA, and a
sample of the largest private primary insurers and reinsurers in the
United States, Europe, and Bermuda. The companies we interviewed
represent about 45 percent of the total domestic insurance market but
should not be generalized to represent all insurance companies. We also
interviewed officials from catastrophe modeling firms, insurance
industry associations, the National Association of Insurance
Commissioners (NAIC),[Footnote 3] and universities to provide
additional context for respondents' statements. To supplement these
interviews, we reviewed documentation of federal agencies' risk
management practices, studies by subject area experts, industry
reports, insurance company documents, and previous GAO reports. We
performed our work between February 2006 and January 2007 in accordance
with generally accepted government auditing standards. A more extensive
discussion of our scope and methodology appears in appendix I.
Results in Brief:
Assessments by the National Academy of Sciences (NAS) and the
Intergovernmental Panel on Climate Change (IPCC), a leading source for
international climate expertise, report that the effects of climate
change on weather-related events and--by extension--weather-related
losses could be substantial. IPCC reports that global mean temperatures
increased by 0.74 degrees Celsius over the last 100 years and are
projected to continue to rise over the next century. Although
temperatures have varied throughout history due to natural processes,
such as changes in the Earth's orbit and volcanic eruptions, the IPCC
and NAS report that the observed temperature increase during the
twentieth century cannot be explained by natural variability alone but
is largely attributable to human activities. Warmer surface
temperatures are linked to global-scale oceanographic, meteorological,
and biological changes. For example, as the earth warms, more water
evaporates from oceans and other sources, eventually falling as rain or
snow. Key assessments that rely on both observational data and computer
models have reported that warmer temperatures are expected to increase
the frequency and severity of damaging extreme weather-related events
(such as flooding or drought), although the timing, magnitude, and
duration of these changes are as yet undetermined. Further research on
the effect of increasing temperature on weather events is ongoing. Of
particular note, the IPCC is expected to release its fourth assessment
of the state of climate science throughout 2007, and the Climate Change
Science Program is currently assessing potential changes in the
frequency or intensity of weather-related events specific to North
America in a report scheduled for release in 2008.
Taken together, private and federal insurers paid more than $320
billion in claims on weather-related losses from 1980 through 2005. In
constant dollars, private insurers paid the largest part of the claims
during this period, $243.5 billion (about 76 percent); followed by
federal crop insurance, $43.6 billion (about 14 percent); and federal
flood insurance, $34.1 billion (about 11 percent). Claims varied
significantly from year to year--largely due to the incidence and
effects of catastrophic weather events such as hurricanes and droughts-
-but generally increased during this period. In particular, the years
with the largest insured losses were generally associated with major
hurricanes, which comprised well over one-third of all weather-related
losses since 1980. The growth in population in hazard-prone areas, and
resulting real estate development and increasing real estate values,
have increased federal and private insurers' exposure, and have helped
to explain the increase in losses. In particular, heavily-populated
areas along the Northeast, Southeast, and Texas coasts have among the
highest value of insured properties in the United States and face the
highest likelihood of major hurricanes. Due to these and other factors,
federal insurers' exposures have grown substantially. Since 1980,
NFIP's exposure has quadrupled, nearing $1 trillion, and program
expansion has increased FCIC's exposure nearly 26-fold to $44 billion.
These escalating exposures to catastrophic weather events are leaving
the federal government at increased financial risk. FCIC officials told
us, for example, that if the widespread Midwest floods of 1993 were to
occur today, losses would be five times greater.
While both major private and federal insurers are exposed to increases
in the frequency or severity of weather-related events associated with
climate change, the two sectors are responding in different ways. Using
computer-based catastrophe models, many major private insurers are
incorporating some near-term elements of climate change into their risk
management practices. One consequence is that, as these insurers seek
to limit their own catastrophic risk exposure, they are transferring
some of it to policyholders and to the public sector. In addition, some
private insurers are approaching climate change at a strategic level by
publishing reports outlining the potential industry-wide impacts and
strategies to proactively address the issue. Federal insurance
programs, on the other hand, have done little to develop the kind of
information needed to understand the programs' long-term exposure to
climate change for a variety of reasons. The federal insurance programs
are not oriented toward earning profits like private insurers but
rather toward increasing participation among eligible parties.
Consequently, neither program has had reason to develop information on
their long-term exposure to the fiscal risks associated with climate
change.
We acknowledge the different mandate and operating environment in which
the major federal insurance programs operate, but we believe that
better information about the federal government's exposure to potential
changes in weather-related risk would help the Congress identify and
manage this emerging high-risk area--one which may not constitute an
immediate crisis, but which does have significant implications for the
nation's growing fiscal imbalance. Accordingly, GAO is recommending
that the Secretary of Agriculture and the Secretary of Homeland
Security direct the Under Secretary for Farm and Foreign Agricultural
Services and the Under Secretary of Homeland Security for Emergency
Preparedness to analyze the potential long-term fiscal implications of
climate change for the FCIC and the NFIP, respectively, and report
their findings to the Congress.
In commenting on a draft of this report, both the Departments of
Agriculture (USDA) and Homeland Security (DHS) agreed with our
recommendation, and USDA commented on the presentation of several
findings in the draft. The Department of Commerce neither agreed nor
disagreed with the report's findings, but instead commented on the
presentation of several issues in the draft and offered technical
comments which we incorporated into this report as appropriate. The
Department of Energy elected not to provide comments on the draft.
Background:
Insurance is a mechanism for spreading risk over time, across large
geographical areas, and among industries and individuals. While
insurers assume some financial risk when they write policies, they
employ various strategies to manage risk so that they earn profits,
limit potential financial exposures, and build capital needed to pay
claims.[Footnote 4] For example, they charge premiums for coverage and
establish underwriting standards, such as refusing to insure customers
who pose unacceptable levels of risk, or limiting coverage in
particular geographic areas. Insurance companies may also purchase
reinsurance to cover specific portions of their financial risk.
Reinsurers use similar strategies to limit their risks, including
charging premiums, establishing underwriting standards, and maintaining
close, long-term business relationships with certain insurers.
Both insurers and reinsurers must also predict the frequency and
severity of insured losses with some reliability to best manage
financial risk.[Footnote 5] In some cases, these losses may be fairly
predictable. For example, the incidence of most automobile insurance
claims is predictable, and losses generally do not occur to large
numbers of policyholders at the same time. However, some infrequent
weather-related events--hurricanes, for example--are so severe that
they pose unique challenges for insurers and reinsurers. Commonly
referred to as catastrophic or extreme events, the unpredictability and
sheer size of these events--both in terms of geography and number of
insured parties affected--have the potential to overwhelm insurers' and
reinsurers' capacity to pay claims. Catastrophic events may affect many
households, businesses, and public infrastructure across large areas,
resulting in substantial losses that deplete insurers' and reinsurers'
capital.
Given the higher levels of capital that reinsurers must hold to address
catastrophic events, reinsurers generally charge higher premiums and
restrict coverage for such events. Further, in the wake of catastrophic
events, reinsurers and insurers may sharply increase premiums to
rebuild capital reserves and may significantly restrict insurance and
reinsurance coverage to limit exposure to similar events in the future.
Under certain circumstances, the private sector may determine that a
risk is uninsurable. For example, while homeowner insurance policies
typically cover damage and losses from fire and other perils, they
usually do not cover flood damage because private insurance companies
are largely unwilling to bear the financial risks associated with its
potentially catastrophic impact. In other instances, the private sector
may be willing to insure a risk, but at rates that are not affordable
to many property owners. Without insurance, affected property owners
must rely on their own resources or seek out disaster assistance from
local, state, and federal sources.
In situations where the private sector will not insure a particular
type of risk, the public sector may create markets to ensure the
availability of insurance. For example, several states have established
Fair Access to Insurance Requirements (FAIR) plans, which pool
resources from insurers doing business in the state to make property
insurance available to property owners who cannot obtain coverage in
the private insurance market, or cannot do so at an affordable rate. In
addition, six southern states have established windstorm insurance
pools that pool resources from private insurers to make insurance
available to property owners who cannot obtain it in the private
insurance market.
Similarly, at the federal level, the Congress established the NFIP and
the FCIC to provide coverage where voluntary markets do not
exist.[Footnote 6] The Congress established the NFIP in 1968, partly to
provide an alternative to disaster assistance for flood damage.
Participating communities are required to adopt and enforce floodplain
management regulations, thereby reducing the risks of flooding and the
costs of repairing flood damage. FEMA, within the Department of
Homeland Security, is responsible for, among other things, oversight
and management of the NFIP. Under the program, the federal government
assumes the liability for covered losses and sets rates and coverage
limitations.
The Congress established the FCIC in 1938 to temper the economic impact
of the Great Depression and the weather effects of the dust bowl. In
1980, the Congress expanded the program to provide an alternative to
disaster assistance for farmers that suffer financial losses when crops
are damaged by droughts, floods, or other natural disasters. Farmers'
participation is voluntary, but the federal government encourages it by
subsidizing their insurance premiums. USDA's RMA is responsible for
administering the crop insurance program, including issuing new
insurance products and expanding existing insurance products to new
geographic regions. RMA administers the program in partnership with
private insurance companies, which share a percentage of the risk of
loss or the opportunity for gain associated with each insurance policy
written.
Climate Change May Increase Losses by Altering the Frequency or
Severity of Weather-Related Events:
Global temperatures have increased in the last 100 years and are
projected to continue to rise over the next century. Using
observational data and computer modeling, climatologists and other
scientists are assessing the likely effects of temperature rise
associated with climate change on precipitation patterns and on the
frequency and severity of weather-related events. The key scientific
assessments we reviewed generally found that warmer temperatures are
expected to alter the frequency or severity of damaging weather-related
events, such as flooding or drought, although the timing, magnitude,
and duration of these changes are as yet undetermined. Additional
research on the effect of increasing temperature on weather events is
expected in the near future. Nevertheless, research suggests that the
potential effects of climate change on damaging weather-related events
could be significant.
Warming Temperatures Are Expected to Alter the Frequency and Severity
of Damaging Extreme Weather-Related Events:
We reviewed the reports released by IPCC, NAS, and the federal Climate
Change Science Program (CCSP) that are shown in figure 1.[Footnote 7]
These leading scientific bodies report that the Earth warmed during the
twentieth century--0.74 degrees Celsius from 1906 to 2005 according to
a recent IPCC report--and is projected to continue to warm for the
foreseeable future.[Footnote 8] IPCC, NAS, CCSP, and other scientific
bodies report that this increase in temperature cannot be explained by
natural variation alone. IPCC's 2001 assessment of the impact of
increasing temperatures on extreme weather events found that it was
likely the frequency and severity of several types of events will
increase as greenhouse gas emissions continue.[Footnote 9]
Figure 1: Time Line of Key Scientific Assessments:
[See PDF for image]
Source: GAO.
[End of figure]
Average Global Temperatures Have Increased and Are Expected to Continue
to Rise:
The earth's climate system is driven by energy from the sun and is
maintained by complex interactions between the atmosphere, the oceans,
and the reflectivity of the earth's surface, among other factors. Upon
reaching the earth, the sun's energy is either reflected back into
space, or is absorbed by the earth and is subsequently reemitted.
However, certain gases in the earth's atmosphere--such as carbon
dioxide and methane--act like the glass in a greenhouse to trap some of
the sun's energy and prevent it from returning to space. While these
gases play an important part in maintaining life on earth, their
accumulation in the atmosphere can significantly increase global
temperatures.
The earth warmed by roughly 0.74 degrees Celsius over the past 100
years, and is projected to continue warming for the foreseeable future.
While temperatures have varied throughout history, triggered by natural
factors such as volcanic eruptions or changes in the earth's orbit, the
key scientific assessments we reviewed have generally concluded that
the observed increase in temperature in the past 100 years cannot be
explained by natural variability alone. In recent years, major
scientific bodies such as the IPCC, NAS, and the Royal Academy (the
United Kingdom's national academy of science) have concluded that human
activities, including the combustion of fossil fuels, industrial and
agriculture processes, landfills, and some land use changes, are
significantly increasing the concentrations of greenhouse gases and, in
turn, global temperatures.
Although climate models produce varying estimates of the extent of
future changes in temperature, NAS and other scientific organizations
have concluded that available evidence points toward continued global
temperature rise. Assuming continued growth in atmospheric
concentration of greenhouse gases, the latest assessment of computer
climate models projects that average global temperatures will warm by
an additional 1.8 to 4.0 degrees Celsius during the next
century.[Footnote 10]
Some scientists have questioned the significance of the earth's present
temperature rise relative to past fluctuations. To address this issue,
the NAS recently assessed the scientific community's efforts to
reconstruct temperatures of the past 2,000 years and place the earth's
current warming in an historical context.[Footnote 11] Based on its
review, the NAS concluded with a high level of confidence that global
mean surface temperature was warmer during the last few decades of the
twentieth century than during any comparable period during the
preceding 400 years. Moreover, NAS cited evidence that temperatures at
many, but not all, individual locations were higher during the past 25
years than any period of comparable length over the past 1,100 years.
IPCC Expects Continued Warming to Alter Frequency and Severity of
Damaging Extreme Weather-Related Events:
Determining the precise nature and extent of the relationship between
average global temperatures and weather-related events is an
exceedingly challenging task. Several key assessments of the state of
this science have addressed the large body of work on this topic. Using
observational data and computer models, scientists are examining the
effects of rising temperatures on precipitation patterns and the
frequency and severity of extreme weather-related events. The
complexity of weather systems, together with the limited statistical
precision of projections of the extent of future temperature change,
often produces different model results, and the results themselves
represent a range of potential future conditions.
Nonetheless, a key assessment of climate model projections indicates
that an increase is likely in the frequency or severity of damaging
extreme weather-related events. In 2001, the IPCC, a leading scientific
authority on climate science, released its Third Assessment Report,
which assessed the state of knowledge of, among other things, the
potential for global changes in extreme weather-related events. The
IPCC described the relationship between temperatures, precipitation,
and weather-related events. Increased global mean surface temperatures
are linked to global-scale oceanographic, meteorological, and
biological changes. For example, as the earth warms, more water
evaporates from oceans or lakes, eventually falling as rain or snow.
IPCC reported that permafrost is thawing, and the extent of sea ice,
snow cover, and mountain glaciers are generally shrinking. The IPCC
also noted that global sea level rose between 0.1 and 0.2 meters during
the twentieth century through thermal expansion of seawater and
widespread loss of land ice, and that this sea level rise could
increase the magnitude of hurricane storm surge in some areas. Warming
is expected to change rainfall patterns, partly because warmer air
holds more moisture.
Based on model projections and expert judgment,[Footnote 12] the IPCC
reported that future increases in the earth's temperature are likely to
increase the frequency and severity of many damaging extreme weather-
related events (summarized in table 1). For instance, IPCC reported
that increased drought is likely across many regions of the globe,
including the U.S. Great Plains. Also, IPCC concluded that the
intensity of precipitation events is very likely to increase across
almost all regions of the globe and that heavy precipitation events are
expected to become more frequent. Compared with projected temperature
increases, changes in the frequency and severity of extreme events can
occur relatively rapidly, according to the IPCC.
Table 1: Selected IPCC Estimates of Confidence in Projected Changes in
Weather-Related Events:
Weather-related event: Higher maximum temperatures and more hot days
over nearly all land areas;
Confidence in projected future changes: Very likely.
Weather-related event: Higher minimum temperatures and fewer cold and
frost days over nearly all land areas;
Confidence in projected future changes: Very likely.
Weather-related event: More intense precipitation events;
Confidence in projected future changes: Very likely.
Weather-related event: Increased summer drying and associated risks of
drought;
Confidence in projected future changes: Likely[A].
Weather-related event: Increase in hurricane peak wind intensities;
Confidence in projected future changes: Likely[B].
Weather-related event: Increase in hurricane average and peak
precipitation intensities;
Confidence in projected future changes: Likely.
Source: IPCC, Climate Change 2001: The Scientific Basis, 2001.
[A] Projections for most midlatitude continental interiors. IPCC found
a lack of consistent projections in other regions.
[B] IPCC reported that changes in the regional distribution of
hurricanes are possible but have not been established.
[End of table]
Much research has been done since the IPCC's Third Assessment Report,
but there has not been a similarly rigorous assessment of what is known
with regard to temperature increase, precipitation, and weather-related
events for the United States.[Footnote 13] However, significant
assessments will be completed in the near future. In particular, the
IPCC is expected to release its Fourth Assessment Report throughout
2007.
While we were completing our review, the IPCC released a summary of the
first of three components of its Fourth Assessment Report, which builds
upon past IPCC assessments and incorporates new findings from the
physical science research since the Third Assessment Report. The
summary reports higher confidence in projected patterns of warming and
other regional-scale features, including changes in wind patterns,
precipitation, and some aspects of extreme events. In particular, the
summary reports that it is very likely that hot extremes, heat waves,
and heavy precipitation events will continue to become more frequent.
Moreover, based on a range of models, IPCC's summary states that it is
likely that future tropical cyclones (typhoons and hurricanes) will
become more intense, with larger peak wind speeds and more heavy
precipitation associated with ongoing increases in tropical sea surface
temperatures. IPCC reports less confidence in projections of a global
decrease in the number of tropical cyclones, and that the apparent
increase in the proportion of very intense storms since 1970 in some
regions is much larger than simulated by current models for that
period. The full first component report was not publicly released prior
to the issuance of our report and is expected some time after May 2007.
The other two components of the Fourth Assessment Report will cover
impacts, adaptation, and vulnerability, and mitigation. These reports
are expected to assess, among other things, key vulnerabilities and
risks from climate change, including changes in extreme events.
Additionally, the IPCC has committed to producing a capping report that
is intended to synthesize and integrate material contained in the
forthcoming reports, as well as other IPCC products.
In addition to the IPCC's work, CCSP is assessing potential changes in
the frequency or intensity of weather-related events specific to North
America in a report scheduled for release in 2008. According to a
National Oceanic and Atmospheric Administration (NOAA) official and
agency documents, the report will focus on weather extremes that have a
significant societal impact, such as extreme cold or heat spells,
tropical and extra-tropical storms, and droughts. Importantly,
officials have said the report will provide an assessment of the
observed changes in weather and climate extremes, as well as future
projections.
More Frequent or More Severe Extreme Weather-Related Events Could
Significantly Increase Insured Losses:
Extreme weather-related events impact communities and economic activity
by damaging homes and vehicles (e.g., see fig. 2), interrupting
electrical service and business operations, or destroying crops. IPCC
reported that the insurance industry--especially the property and
casualty segment--are sensitive to the effects of weather-related
events. This was highlighted in the Department of Commerce's comments
on a draft of this report, which observed that altering either the
frequency or severity of high impact extreme weather-related events
could result in a significant increase in the risk posed to an insurer.
For example, the agency said that what had been considered a 500-year
event (i.e., its probability of occurring in a given year is 1 in 500)
could shift under climate change to become a 100-year event (i.e., its
probability of occurring in a given year is 1 in 100). Consequently,
more frequent or more severe events have a greater potential for damage
and, in turn, insured losses. As an official from Aon Re Australia, a
large global reinsurer, reported, "The most obvious impact of climate
change on the insurance sector will be the increase in insured property
losses from extreme weather events."[Footnote 14]
Figure 2: July 1993 Flood Damage at Chesterfield Airport in St. Louis,
Missouri:
[See PDF for image]
Source: FEMA.
Note: According to FEMA, the depth of the floodwaters underscores the
extent of the damage caused by the 1993 Midwest flood. A total of 534
counties in nine states were declared for federal disaster aid.
[End of figure]
Notably, the economic damages associated with some extreme weather-
related events could increase at a greater rate in comparison with
changes in the events themselves. Seemingly small changes in the
characteristics of certain weather-related events can lead to
substantial increases in damage. For example, recent work on hurricanes
by researchers at the University of Colorado, the National Weather
Service, and other institutions examined losses associated with
hurricanes that made landfall in the United States since 1900.[Footnote
15] Holding constant the increased population and development in
coastal counties during this period, the study compared the economic
damage of stronger storms with weaker storms, based on the Saffir-
Simpson Hurricane Scale.[Footnote 16] The researchers found that
stronger storms have caused many times more economic damages than
weaker storms, as shown in figure 3. These findings are consistent with
other independent analyses conducted by insurers and catastrophe
modelers.
Figure 3: Economic Damages by Hurricane Category for U.S. Hurricanes
Making Landfall, 1900-2005:
[See PDF for image]
Source: GAO adaptation of Pielke et al. data.
Note: Value of each bar compares the median economic damage associated
with hurricanes of that Saffir-Simpson category with the median
economic damage of Category One storms. Of the 158 hurricanes reviewed,
only three were Category Five.
[End of figure]
Moreover, public reports from several of the world's largest
reinsurance companies and brokers underscore the potential for
substantially increased losses. These reports note that, in addition to
greater losses in absolute terms, the potential for greater variability
in weather- related events could significantly enhance the volatility
of losses.
Insured Weather-Related Losses Have Been Sizeable, and Federal
Insurers' Exposure Has Grown Significantly:
Taken together, insurers paid more than $320 billion in claims for
weather-related losses between 1980 and 2005.[Footnote 17] Claims
varied significantly from year to year--largely due to the effects of
catastrophic weather events such as hurricanes and droughts--but
generally increased during this period. The growth in population in
hazard-prone areas, and consequent real estate development and
increasing real estate values, have generally increased insurers'
exposure to weather-related events and help to explain their increased
losses. Due to these and other factors, the federal insurance programs'
liabilities have grown significantly, leaving the federal government
increasingly vulnerable to the financial impacts of extreme events.
Claims Paid on Weather-Related Losses Totaled More Than $320 Billion
between 1980 and 2005:
Based on an examination of loss data from several different sources,
insurers incurred more than $320 billion in weather-related losses from
1980 through 2005 (see fig. 4). Weather-related losses accounted for 88
percent of all property losses paid by insurers during this period. All
other property losses, including those associated with earthquakes and
terrorist events, accounted for the remainder. Weather-related losses
varied significantly from year to year, ranging from just over $2
billion in 1987 to more than $75 billion in 2005.
Figure 4: Annual Weather-and Nonweather-Related Insured Losses:
[See PDF for image]
Source: GAO analysis of PCS, NFIP, and FCIC data.
[End of figure]
Privately-Insured Losses:
Of the $321.2 billion in weather-related loss payments we reviewed,
private insurers paid $243.5 billion--over three-quarters of the
total.[Footnote 18] Figure 5 depicts the breakdown of these payments
among key weather-related events. Of the $243.5 billion paid by private
insurers, hurricanes accounted for $124.6 billion, or slightly more
than half. Wind, tornados, and hail associated with severe
thunderstorms accounted for $77 billion, or nearly one-third of the
private total. Winter storms were associated with $25.1 billion, or
about 10 percent.
Figure 5: Weather-Related Losses Paid by Private Insurers:
[See PDF for image]
Source: GAO analysis of PCS data.
[End of figure]
Federally-Insured Losses:
The two major federal insurance programs--NFIP and FCIC--paid the
remaining $77.7 billion of the $321.2 billion in weather-related loss
payments we reviewed.[Footnote 19] Although the performance of both
NFIP and FCIC is sensitive to weather, the two programs insure
fundamentally different risks and operate in very different ways.
NFIP provides insurance for flood damage to homeowners and commercial
property owners in more than 20,000 communities. Homeowners with
mortgages from federally regulated lenders on property in communities
identified as being in high flood risk areas are required to purchase
flood insurance on their dwellings. Optional, lower cost flood
insurance is also available under the NFIP for properties in areas of
lower flood risk. NFIP offers coverage for both the property and its
contents, which may be purchased separately.
NFIP claims totaled about $34.1 billion, or about 11 percent of all
weather-related insurance claims during this period. As shown in figure
6, NFIP covers only one cause of loss--flooding. Claims averaged about
$1.3 billion per year, but ranged from $75.7 million in 1988 to $16.7
billion in 2005.
Figure 6: Weather-Related Losses Paid by NFIP:
[See PDF for image]
Source: GAO analysis of NFIP data.
[End of figure]
FCIC insures commodities on a crop-by-crop and county-by-county basis
based on farmer demand for coverage and the level of risk associated
with the crop in a given region. Over 100 crops are covered by the
program. Major crops, such as grains, are covered in almost every
county where they are grown, and specialty crops, such as fruit, are
covered only in some areas. Participating farmers can purchase
different types of crop insurance, including yield and revenue
insurance, and at different levels. For yield insurance, participating
farmers select the percentage of yield of a covered crop to be insured
and the percentage of the commodity price received as payment if the
producer's losses exceed the selected threshold. Revenue insurance pays
if actual revenue falls short of an assigned target level regardless of
whether the shortfall was due to low yield or low commodity market
prices.
Since 1980, FCIC claims totaled $43.6 billion, or about 14 percent of
all weather-related claims during this period. FCIC losses averaged
about $1.7 billion per year, ranging from $531.8 million in 1987 to
$4.2 billion in 2002. Figure 7 shows the three causes of loss--drought,
excess moisture, and hail--that accounted for more than three-quarters
of crop insurance claims. In particular, drought accounted for $18.6
billion in losses, or more than 40 percent of all insured crop losses.
Excess moisture totaled $11.2 billion, followed by hail with total
claims of $4.2 billion. The remaining $9.6 billion in claims was spread
among 27 different causes of loss, including frost and tornados.
Figure 7: Weather-Related Losses Paid by FCIC:
[See PDF for image]
Source: GAO analysis of FCIC data.
[End of figure]
Insured Losses Understate Total Economic Damage:
Importantly, the insured loss totals used in our analysis do not
account for all economic damage associated with weather-related
events.[Footnote 20] Specifically, data are not available for several
categories of economic losses, including uninsured, underinsured, and
self-insured losses. As we reported in 2005, FEMA estimates that one-
half to two-thirds of structures in floodplains do not have flood
insurance because the uninsured owners either are unaware that
homeowners insurance does not cover flood damage, or they do not
perceive a serious flood risk.[Footnote 21] Furthermore, industry
analysts estimate that 58 percent of homeowners in the United States
are underinsured--that is, they carry a policy below the replacement
value of their property--by an average of 21 percent.[Footnote 22]
Finally, some individuals and businesses have the means to "self-
insure" their assets by assuming the full risk of any damage.
Various public and private disaster relief organizations provide
assistance to communities and individuals who suffer noninsured
economic losses, although it was beyond the scope of this report to
collect data on these losses. In particular, since 1989, $78.6 billion
in federal disaster assistance funds have been obligated through the
Disaster Relief Fund administered by FEMA, the largest--but not only--
conduit for federal disaster assistance money provided in the wake of
presidentially declared disasters and emergencies.
Overall, according to data obtained from Munich Re, one of the world's
largest reinsurers, the type of insured losses we reviewed account for
no more than about 40 percent of the total losses attributable to
weather-related events.[Footnote 23] NOAA's National Hurricane Center
(NHC) uses a similar proportion to produce the agency's estimates of
total economic damage attributable to hurricanes.[Footnote 24] Although
we did not independently evaluate the reliability of these estimates,
subject area experts we spoke with confirmed that it was the best such
estimate available and is widely used as an approximation of the
relative distribution of losses.
The difficulties we and others faced in accounting for weather-related
losses were the subject of the National Academies' The Impacts of
Natural Disasters: A Framework for Loss Estimation.[Footnote 25]
Reporting how best to account for the costs of natural disasters,
including weather-related events, NAS found that there was no system in
place in either the public or the private sectors to consistently
capture information about the economic impact. Specifically, the NAS
report found no widely accepted framework, formula, or method for
estimating these losses. Moreover, NAS found no comprehensive
clearinghouse for the disaster loss information that is currently
collected. To that end, NAS recommended that the Office of Management
and Budget, in consultation with FEMA and other federal agencies,
develop annual, comprehensive estimates of the payouts for disaster
losses made by federal agencies. Reviewing the status of this
recommendation was beyond the scope of this report. Nevertheless, our
experience with trying to obtain comprehensive information on disaster
costs and losses underscores the NAS findings.
Catastrophic Weather-Related Events Help Explain the Significant Year-
to-Year Variance in Losses:
The largest insured losses in the data we reviewed were associated with
catastrophic weather events. These events have a low probability of
occurrence, but their consequences are severe. Notably, both crop
insurers and other property insurers face the catastrophic risks posed
by extreme events, although the nature of the events for each is very
different. In the case of crop insurance, drought accounted for more
than 40 percent of all insured losses from 1980 to 2005, and the years
with the largest losses were associated with drought. Taken together,
though, hurricanes were the most damaging event experienced by insurers
in the data we reviewed. Although the United States experienced an
average of only two hurricanes per year from 1980 through 2005, weather-
related claims attributable to hurricanes totaled more than 45 percent
of all weather-related insured losses--more than $146 billion.
Moreover, these losses appear to be increasing.
In the data we reviewed, the years with the largest insured losses were
generally associated with major hurricanes, defined as Category Three,
Four, or Five on the Saffir-Simpson Hurricane Scale. Table 2 shows
that, while 29 Category One and Two storms account for nearly $18
billion in losses, the 21 major storms account for over $126 billion in
losses. In fact, claims associated with major hurricanes comprised 40
percent of all weather-related insured losses since 1980.
Table 2: Insured Losses Associated with Hurricanes:
Dollars in thousands.
1980s;
Categories One, Two: $807,422 (11);
Categories Three, Four, Five: $9,905,042 (6);
Total: $10,712,464 (17).
1990s;
Categories One, Two: 9,038,801 (11);
Categories Three, Four, Five: 29,099,303 (8);
Total: 38,138,104 (19).
2000s;
Categories One, Two: 8,071,619 (7);
Categories Three, Four, Five: 89,210,093 (7);
Total: 97,281,712 (14).
Total;
Categories One, Two: $17,917,842 (29);
Categories Three, Four, Five: $128,214,438 (21);
Total: $146,132,280 (50).
Sources: GAO analysis of PCS and NFIP data; NOAA (hurricane intensity
classification).
Note: Totals do not include crop losses associated with hurricanes.
Number of hurricanes associated with losses is included in parentheses.
Hurricane classification was based on peak intensity at landfall.
[End of table]
Importantly, hurricane severity is only one factor in determining the
size of a particular loss--the location affected by the hurricane is
also important. Generally, the more densely populated an area, the
greater the extent of economic activity and accumulated value of the
building stock. For instance, several studies have reviewed the
economic impact of Hurricane Andrew, which tracked over Florida in
1992, in light of the dramatic real estate development that has
occurred in the meantime. Researchers have normalized losses associated
with the storm to account for societal changes by holding constant the
value of building materials, real estate, and other factors so that the
storm's impact could be adjusted to reflect contemporary
conditions.[Footnote 26] Hurricane Andrew, which resulted in roughly
$25 billion in total economic losses in 1992, would have resulted in
more than twice that amount--$55 billion--were it to have occurred in
2005, given current asset values.
Several recent studies have commented on the apparent increases in
hurricane losses during this time period, and weather-related disaster
losses generally, with markedly different interpretations. Some argue
that loss trends are largely explained by changes in societal and
economic factors, such as population density, cost of building
materials, and the structure of insurance policies.[Footnote 27] Others
argue that increases in losses have been driven by changes in
climate.[Footnote 28]
To address this issue, Munich Re and the University of Colorado's
Center for Science and Technology Policy Research jointly convened a
workshop in Germany in May 2006 to assess factors leading to increasing
weather-related loss trends.[Footnote 29] The workshop brought together
a diverse group of international experts in the fields of climatology
and disaster research. Among other things, the workshop sought to
determine whether the costs of weather-related events were increasing
and what factors account for increasing costs in recent decades.
Workshop participants reached consensus on several points, including
that analyses of long-term records of disaster losses indicate that
societal change and economic development are the principal factors
explaining observed increases in weather-related losses.[Footnote 30]
However, participants also agreed that changing patterns of extreme
events are drivers for recent increases in losses and that additional
increases in losses are likely given IPCC's projected increase in the
frequency or severity of weather-related events.
Value at Risk in Federal Insurers' Portfolios Increased Significantly
between 1980 and 2005:
The growth in population in hazard-prone areas, and consequent real
estate development and increasing real estate values, are leaving the
nation increasingly exposed to higher insured losses. The close
relationship between the value of the resource exposed to weather-
related losses and the amount of damage incurred may have ominous
implications for a nation experiencing rapid growth in some of its most
disaster-prone areas. We reported in 2002 that the insurance industry
faces potentially significant financial exposure due to natural
catastrophes.[Footnote 31] Heavily populated areas along the Northeast,
Southeast, and Texas coasts have among the highest value of insured
properties in the United States and face the highest likelihood of
major hurricanes. According to insurance industry estimates, a large
hurricane in Miami could cause up to $110 billion in insured losses
with total losses as high as $225 billion. Several states--including
Florida, California, and Texas--have established programs to help
ensure that coverage is available in areas particularly prone to these
events.[Footnote 32]
AIR Worldwide, a leading catastrophe modeling firm, recently reported
that insured losses should be expected to double roughly every 10 years
because of increases in construction costs, increases in the number of
structures, and changes in their characteristics. AIR's research
estimates that, because of exposure growth, probable maximum
catastrophe loss grew in constant dollars from $60 billion in 1995 to
$110 billion in 2005, and it will likely grow to over $200 billion
during the next 10 years.
Data obtained from both the NFIP and FCIC programs indicate the federal
government has grown markedly more exposed to weather-related losses
regardless of the cause. For example, NFIP data show that the number of
policyholders and the value of the properties insured have both
increased since 1980. Figure 8 shows the growth of NFIP's exposure in
terms of both number of policies and the total coverage. The number of
policies has more than doubled in this time period, from 1.9 million
policies to more than 4.6 million. Moreover, although NFIP limits
coverage to $250,000 for a personal structure and $100,000 for its
contents, and $500,000 of coverage for a business structure and
$500,000 on its contents, more policyholders' homes are approaching (or
exceeding) these coverage limits. Accordingly, the total value covered
by the program increased fourfold in constant dollars during this time
from about $207 billion to $875 billion in 2005.
Figure 8: NFIP Policies and Total Coverage:
[See PDF for image]
Source: GAO analysis of NFIP data.
[End of figure]
Similarly, RMA data show that FCIC has effectively increased its
exposure base 26-fold during this period (in constant dollars). In
particular, the program has significantly expanded the scope of crops
covered and increased participation. Figure 9 shows the growth in FCIC
exposure since 1980.[Footnote 33]
Figure 9: FCIC Total Coverage:
[See PDF for image]
Source: GAO analysis of FCIC data.
[End of figure]
A senior RMA official told us that the main implication of FCIC's
growth is that the magnitude of potential claims, in absolute terms, is
much greater today than in the past. For example, if the Midwest floods
of 1993 were to occur today, losses would be five times greater than
the $2 billion paid in 1993, according to RMA officials.
Major Private and Public Insurers Differ in How They Manage
Catastrophic Risks Associated with Climate Change:
Although the relative contribution of event intensity versus societal
factors in explaining the rising losses associated with weather-related
events is still under investigation, both major private and federal
insurers are exposed to increases in the frequency or severity of
weather-related events associated with climate change. Nonetheless,
major private and federal insurers are responding to this prospect
differently. Many large private insurers are incorporating some
elements of near-term climate change into their risk management
practices. Furthermore, some of the world's largest insurers have also
taken a long-term strategic approach toward changes in climate. On the
other hand, for a variety of reasons, the federal insurance programs
have done little to develop the kind of information needed to
understand the programs' long-term exposure to climate change. We
acknowledge the different mandate and operating environment in which
the major federal insurance programs operate but believe that better
information about the federal government's exposure to potential
changes in weather-related risk would help the Congress identify and
manage this emerging high-risk area; one which may not constitute an
immediate crisis but which may pose an important longer term threat to
the nation's welfare.
Major Private Insurers Prospectively Manage Potential Increases in
Catastrophic Risk Associated with Climate Change:
Extreme weather events pose a unique financial threat to private
insurers' financial success because a single event can cause insolvency
or a precipitous drop in earnings, liquidation of assets to meet cash
needs, or a downgrade in the market ratings used to evaluate the
soundness of companies in the industry. To prevent these disruptions,
the American Academy of Actuaries (AAA)--the professional society that
establishes, maintains, and enforces standards of qualification,
practice, and conduct for actuaries in the United States--has outlined
a five-step process for private insurers to follow to manage their
catastrophic risk. These steps include the following:
* identifying catastrophic risk appetite by determining the maximum
potential loss they are willing to accept;
* measuring catastrophic exposure by determining how vulnerable their
total portfolio is to loss, both in absolute terms and relative to the
company's risk management goals;
* pricing for catastrophic exposure by setting rates to collect
sufficient premiums to cover their expected catastrophic loss and other
expenses;
* controlling catastrophic exposure by reducing their policies in areas
where they have too much exposure, or transferring risk using
reinsurance or other mechanisms; and:
* evaluating their ability to pay claims by determining the sufficiency
of their financial resources to cover claims in the event of a
catastrophe.
Additionally, insurers monitor their exposure to catastrophic weather-
related risk using sophisticated computer models called "catastrophe
models."[Footnote 34] AAA emphasizes the shortcomings of estimating
future catastrophic risk by extrapolating solely from historical losses
and endorses catastrophe models as a more rigorous approach.[Footnote
35] Catastrophe models incorporate the underlying trends and factors in
weather phenomena and current demographic, financial, and scientific
data to estimate losses associated with various weather-related events.
According to an industry representative, catastrophe models assess a
wider range of possible events than the historical loss record alone.
These models simulate losses from thousands of potential catastrophic
weather-related events that insurers use to better assess and control
their exposure and inform pricing and capital management decisions.
Figure 10 illustrates the difference between estimating future
catastrophic losses using historical data versus catastrophe models.
Figure 10: Modeling Potential Catastrophe Losses:
[See PDF for image]
Source: Adapted from the American Academy of Actuaries and Towers
Perrin.
[End of figure]
To determine what major private insurers are doing to estimate and
prepare for risks associated with potential changes in climate arising
from natural or human factors, we contacted 11 of the largest private
insurers operating in the U.S. property casualty insurance market.
Representatives from each of the 11 major insurers we interviewed told
us they use catastrophe models that incorporate a near-term higher
frequency and intensity of hurricanes. Of the 11 private insurers, 6
specifically attributed the higher frequency and intensity of
hurricanes to the Atlantic Multidecadal Oscillation, which--according
to NOAA--is a 20-to 40-year climatic cycle of fluctuating temperatures
in the north Atlantic Ocean. The remaining 5 insurers did not elaborate
on the elements of climate change driving the differences in hurricane
characteristics.
Industry reports indicate that insurance companies' perception of
increased risk from hurricanes has prompted them to reduce their near-
term catastrophic exposure, in both reinsurance and primary insurance
coverage along the Gulf Coast and eastern seaboard. For example, a
recent industry analysis from a leading insurance broker reported that
reinsurance coverage is substantially limited in the southeastern
United States and that reinsurance prices have more than doubled from
2005 to 2006, following a record-setting hurricane season.[Footnote 36]
According to the Insurance Information Institute, a leading source of
information about the insurance industry, primary insurance companies
have also raised prices in coastal states to cover rising reinsurance
costs.[Footnote 37] Additionally, a recent report co-authored by a
major international insurance company cites several examples of large
primary insurers either limiting coverage or withdrawing from
vulnerable areas such as Florida,[Footnote 38] the Gulf Coast, and Long
Island.[Footnote 39]
As private insurers limit their exposure, catastrophic risk is
transferred to policyholders and the public sector. Insurance companies
transfer risk to policyholders by increasing premiums and deductibles,
or by setting lower coverage limits for policies. Insurers can also
transfer risk to policyholders by passing along the mandatory
participation costs of state-sponsored insurance plans.[Footnote 40]
For example, after the 2004 hurricane season, insurers assessed a
surcharge of about 7 percent to every policyholder in Florida to recoup
the cost of insurers' participation in the state-sponsored wind
insurance plan. The public sector assumes management of weather-related
risk at the local, state, and national level by providing disaster
relief and recovery, developing mitigation projects, appropriating
funds and, ultimately, providing insurance programs when private
insurance markets are not sufficient or do not exist.
In addition to managing their aggregate exposure on a near-term basis,
some of the world's largest insurers have also taken a long-term
strategic approach to changes in catastrophic risk. For example, major
insurance and reinsurance companies, such as Allianz, Swiss Re, Munich
Re, and Lloyds of London, have published reports that advocate
increased industry awareness of the potential risks of climate change
and outline strategies to address the issue proactively. Moreover, 6 of
the 11 private insurers we interviewed provided one or more additional
activities they have undertaken when asked if their company addresses
changes in climate through their weather-related risk management
processes. These activities include monitoring scientific research (4
insurers), simulating the impact of a large loss event on their
portfolios (3 insurers), and educating others in the industry about the
risks of climate change (3 insurers), among others.
Furthermore, recent research on insurers' activities to address climate
change outlines several other actions that private sector companies are
taking, such as developing specialized policies and new products,
evaluating risks to company stock investments, and disclosing to
shareholders information about company-specific risks due to climate
change.[Footnote 41] Additionally, concern over the potential impacts
of climate change on the availability and affordability of private
insurance has led state insurance regulators to establish a task force
to formally address the issue. The report, issued by the NAIC, is
expected to be published in the summer of 2007.
Major Federal Insurers Have Taken Little Action to Prospectively Assess
Potential Increases in Catastrophic Risk Associated with Climate
Change:
The goals of the major federal insurance programs are fundamentally
different from those of private insurers. Specifically, whereas private
insurers stress the financial success of their business operations, the
statutes governing the NFIP and FCIC promote affordable coverage and
broad participation by individuals at risk. Although both programs
manage risk within their statutory guidelines, unlike the private
sector, neither program is required to limit its catastrophic risk
strictly within the programs' ability to pay claims on an annual basis.
One important implication of the federal insurers' risk management
approach is that they each have little reason to develop information on
their long-term exposure to the potential risk of increased low-
frequency, high-severity weather events associated with climate change.
The statutes governing the NFIP and FCIC promote broad participation
over financial self-sufficiency in two ways: (1) by offering discounted
or subsidized premiums to encourage participation and (2) by making
additional funds available during high-loss years.[Footnote 42] For
example, discounted insurance premiums are available under the NFIP for
some older homes situated within high flood risk areas where insurance
would otherwise have been prohibitively expensive. FEMA is also
authorized to borrow additional federal funds for the NFIP on an as-
needed basis, subject to statutory limits, to cope with
catastrophes.[Footnote 43] One effect has been that the NFIP's exposure
has expanded well beyond the ability to pay claims in high-loss years.
Similar to the discounted premiums offered by the NFIP, the FCIC's
subsidized premiums are designed to make crop insurance available and
affordable to as many participants as possible. For example, the FCIC
is mandated to provide fully subsidized catastrophic coverage for
producers in exchange for a minimal administrative fee, as well as
partial subsidies for additional levels of coverage. Also like the
NFIP, the FCIC is authorized to use additional federal funds on an as-
needed basis during high-loss years--although, unlike the NFIP, the
FCIC is not required to reimburse those additional funds.
Unlike the private sector, the NFIP and the FCIC can use additional
federal funds, and so neither program is required to assess and limit
its catastrophic risk strictly within its ability to pay claims on an
annual basis. Instead, each program manages its risk to the extent
possible, within the context of its broader purposes, in accordance
with its authorizing statutes and implementing regulations.[Footnote
44] For example, the FCIC uses coverage limits, exclusions, and premium
rates to meet their statutory goal of a long-term loss ratio no greater
than 1.075--including premium subsidies.[Footnote 45] Although the
program has experienced high-loss years that required additional
federal funds, over time, these high-loss years have been offset by low-
loss years, which have allowed the program to meet its goal and build
reserves.[Footnote 46]
By developing a goal to generate sufficient revenue to pay for an
average loss year, the NFIP has also been able to generate a surplus in
low-loss years despite borrowing funds in high-loss years. In the past,
the program has been able to repay borrowed funds with interest to the
Department of the Treasury, however, it is unlikely FEMA will be able
to repay the nearly $21 billion borrowed following the 2005 hurricane
season based on the program's current premium income.
Although neither program faces the potential of financial ruin like the
private sector, both programs have occasionally attempted to estimate
their aggregate losses from potential catastrophic events. For example,
FCIC officials stated that they had modeled past events, such as the
1993 Midwest floods, using current participation levels to inform
negotiations with private crop insurers over reinsurance terms. NFIP
and FCIC officials explained that these efforts were informal exercises
and were not performed on a regular basis. FCIC officials also said
they use a hurricane model developed by NOAA to inform pricing
decisions for some commodities such as citrus crops, according to FCIC
officials. However, unlike the catastrophic risk faced by private
insurers, hurricane damages have not been a primary source of crop
insurance claims.
According to NFIP and FCIC officials, their risk management processes
adapt to near-term changes in weather as they affect existing data. As
one NFIP official explained, NFIP is designed to assess and insure
against current--not future--risks. Over time, agency officials stated,
this process has allowed their programs to operate as intended.
However, unlike the private sector, neither program has conducted an
analysis to assess the potential impacts of an increase in the
frequency or severity of weather-related events on their program
operations over the near-or long-term.
Information on Federal Agencies' Long-term Exposure to Catastrophic
Risk Could Better Inform Congressional Decision Making:
While comprehensive information on federal insurers' long-term exposure
to catastrophic risk associated with climate change may not inform the
NFIP's or FCIC's annual operations, it could nonetheless provide
valuable information for the Congress and other policymakers who need
to understand and prepare for fiscal challenges that extend well beyond
the two programs' near-term operational horizons. We have highlighted
the need for this kind of strategic information in recent reports that
have expressed concern about the looming fiscal imbalances facing the
nation. In one report, for example, we observed that, "Our policy
process will be challenged to act with more foresight to take early
action on problems that may not constitute an urgent crisis but pose
important long-term threats to the nation's fiscal, economic, security,
and societal future."[Footnote 47] The prospect of increasing program
exposure, coupled with expected increases in frequency and severity of
weather events associated with climate change, would appear to pose
such a problem.
Agency officials identified several challenges that could complicate
their efforts to assess these impacts at the program level. Both NFIP
and FCIC officials stated there was insufficient scientific information
on projected impacts at the regional and local levels to accurately
assess their impact on the flood and crop insurance programs. However,
members of the insurance industry have analyzed and identified the
potential risks climate change poses, despite similar challenges.
Moreover, as previously discussed, both the IPCC and CCSP are expected
to release significant assessments of the likely effect of increasing
temperatures on weather events in coming months.
The experience of many private insurers, who must proactively respond
to long-term changes in weather-related risk to remain solvent,
suggests the kind of information that might be developed to help
congressional and other policymakers in assessing current and
alternative strategies. Specifically, to help ensure their future
viability, a growing number of private insurers are actively
incorporating the potential for climate change into their strategic
level analyses. In particular, some private insurers have run a variety
of simulation exercises to determine the potential business impact of
an increase in the frequency and severity of weather events. For
example, one insurer simulated the impact of large weather events
occurring simultaneously. A similar analysis could provide the Congress
with valuable information about the potential scale of losses facing
the NFIP and FCIC in coming decades, particularly in light of the
programs' expansion since 1980.
Conclusions:
Recent assessments by leading scientific bodies provide sufficient
cause for concern that climate change may have a broad range of long-
term consequences for the United States and its citizens. While a
number of key uncertainties regarding the timing, location, and
magnitude of impacts remain, climate change has implications for the
fiscal health of the federal government, which already faces other
significant challenges in meeting its long-term fiscal obligations.
NFIP and FCIC are two major federal programs which, as a consequence of
both future climate change and substantial growth in exposure, may see
their losses grow by many billions of dollars in coming decades.
We acknowledge that to carry out their primary missions, these public
insurance programs must focus on the near-term goals of ensuring
affordable coverage for individuals in hazard-prone areas. Nonetheless,
we believe the two programs are uniquely positioned to provide
strategic information on the potential impacts of climate change--
information that would be of value to key decision makers charged with
such a long-term focus. Most notably, in exercising its oversight
responsibilities, the Congress could use such information to examine
whether the current structure and incentives of the federal insurance
programs adequately address the challenges posed by potential increases
in the frequency and severity of catastrophic weather events. While the
precise content of these analyses can be debated, the activities of
many private insurers already suggest a number of strong possibilities
that may be applicable to assessing the potential implications of
climate change on the federal insurance programs.
Recommendation for Executive Action:
We recommend that the Secretary of Agriculture and the Secretary of
Homeland Security direct the Administrator of the Risk Management
Agency and the Under Secretary of Homeland Security for Emergency
Preparedness to analyze the potential long-term implications of climate
change for the Federal Crop Insurance Corporation and the National
Flood Insurance Program, respectively, and report their findings to the
Congress. This analysis should use forthcoming assessments from the
Climate Change Science Program and the Intergovernmental Panel on
Climate Change to establish sound estimates of expected future
conditions. Key components of this analysis may include: (1) realistic
scenarios of future losses under anticipated climatic conditions and
expected exposure levels, including both potential budgetary
implications and consequences for continued program operation and (2)
potential mitigation options that each program might use to reduce
their exposure to loss.
Agency Comments and Our Evaluation:
We provided a draft of this report to the Departments of Agriculture
(USDA), Commerce, Energy, and Homeland Security (DHS) for their review.
DHS agreed via email with the report's recommendation, noting that
conducting an assessment of the impact of climate change beyond FEMA's
current statistical modeling (which is based on historical loss
experience) could be helpful if resources were available to pursue such
an analysis.
USDA also agreed with the report's recommendation, and commented on the
presentation of several findings. (See app. V for the letter from the
Under Secretary for Farm and Foreign Agricultural Services and GAO's
point-by-point response.) In particular, USDA disagreed that it had
thus far taken little action to prospectively assess potential
increases in catastrophic risk associated with climate change. USDA
explained that RMA does assess both the current and long-term exposure
of the crop insurance program to catastrophic weather events, noting
specifically that RMA (1) updates and publishes total program liability
on a weekly basis and (2) estimates expected changes in liability up to
10 years ahead through its baseline projections. We acknowledge these
activities, but believe it is important to note that they are limited
in scope, focusing almost exclusively on retrospective measures of
performance and not on the potential for increasingly frequent and
intense weather-related events. These events, including drought and
heavy precipitation events, are the key events acknowledged by USDA as
posing catastrophic risk to the crop insurance program. Moreover, other
RMA efforts to capture changes in weather-related risk rely on data
reflecting what has been experienced in the past, not on what could be
experienced in the future.
The Department of Commerce neither agreed nor disagreed with the
report's findings, but instead offered several comments on the
presentation of several issues in the draft (particularly the depth in
which several issues are discussed) as well as technical comments. We
have incorporated these comments as appropriate and address them in
detail in appendix VI. Notably, the Department of Commerce underscored
the vulnerability of high-risk coastal development, stating that such
vulnerabilities will only be amplified by climate change-related
increases in the frequency or severity of weather-related events.
Finally, the Department of Energy elected not to provide comments on
the draft.
As agreed with your offices, unless you publicly announce the contents
of this report earlier, we plan no further distribution until 30 days
from the report date. At that time, we will send copies of this report
to the Secretaries of Agriculture, Commerce, Energy, and Homeland
Security, as well as other interested parties. We also will make copies
available to others upon request. In addition, the report will be
available at no charge on the GAO Web site at http://www.gao.gov.
If you or your staff has any questions regarding this report, please
contact me at (202) 512-3841 or stephensonj@gao.gov. Contact points for
our Offices of Congressional Relations and Public Affairs may be found
on the last page of this report. Key contributors are listed in
appendix VII.
Signed by:
John B. Stephenson:
Director, Natural Resources and Environment:
[End of section]
Appendix I: Scope and Methodology:
We were asked us to (1) describe what is known about how climate change
might affect insured and uninsured losses, (2) determine insured losses
incurred by major federal agencies and private insurers and reinsurers
resulting from weather-related events, and (3) determine what major
federal agencies and private insurers and reinsurers are doing to
assess and manage the potential risk of increased losses due to changes
in the frequency and severity of weather-related events associated with
climate change.
Scientific Literature:
To address the first objective, we reviewed and summarized existing
literature from significant policy-oriented scientific assessments from
reputable international and national research organizations including
the Intergovernmental Panel on Climate Change, National Academy of
Sciences, and the multifederal agency U.S. Climate Change Science
Program, as specified in table 3. It was beyond the scope of this
report to independently evaluate the results of these studies.
Table 3: Key Policy-Oriented Scientific Assessments Reviewed by GAO:
Organization: Intergovernmental Panel on Climate Change (IPCC);
Publication:
* Climate Change 2007: The Physical Science Basis, Summary for
Policymakers (2007);
* Climate Change 2001: Synthesis Report (2001);
* Climate Change 2001: The Scientific Basis (2001);
* Climate Change 2001: Impacts, Adaptation & Vulnerability (2001).
Organization: Climate Change Science Program (CCSP);
Publication:
* Temperature Trends in the Lower Atmosphere: Steps for Understanding
and Reconciling Differences, Synthesis and Assessment Product 1.1
(2006).
Organization: National Academy of Sciences (NAS);
Publication:
* Surface Temperature Reconstructions for the Last 2,000 Years (2006);
* Understanding and Responding to Climate Change: Highlights of
National Academies Reports (2006);
* Radiative Forcing of Climate Change: Expanding the Concept and
Addressing Uncertainties (2005);
* From Climate to Weather: Impacts on Society and Economy-Summary of a
Forum, June 28, 2002, Washington, D.C. (2003);
* Understanding Climate Change Feedbacks (2003);
* Abrupt Climate Change: Inevitable Surprises (2002);
* Climate Change Science: An Analysis of Some Key Questions (2001).
Source: GAO.
Note: Publication year follows publication title in parentheses.
[End of table]
Insured Loss Data:
To address the second objective, we analyzed insured loss data from
January 1, 1980, through December 31, 2005, from the Federal Emergency
Management Agency (FEMA) for the National Flood Insurance Program
(NFIP); the Department of Agriculture's Risk Management Agency (RMA)
for the Federal Crop Insurance Corporation (FCIC); and the Property
Claim Services (PCS) for private property insurance. Through electronic
testing and other means, we assessed the reliability of each of the
data sets to determine whether the data were sufficiently reliable for
our purposes. Specifically, we interviewed the sources for each of the
data sets to gather information on how records were collected,
processed, and maintained. Because not all catastrophes are weather-
related, we excluded all events attributable to terrorist acts,
tsunamis, earthquakes, and other nonweather-related losses, based on
discussions with the data provider. To adjust for the general effects
of inflation over time we used the chain-weighted gross domestic
product price index to express dollar amounts in inflation-adjusted
2005 dollars. We reviewed any changes in data collection methodologies
that have occurred over time, and evaluated the effect of any changes
on our ability to report losses. We believe that these data are
sufficiently reliable for the purpose of describing insured losses. We
note, however, that these data likely understate the actual insured
losses.
PCS:
PCS data are estimates of insured losses, or claims paid by private
insurance companies, for catastrophe loss events for the 50 states, as
well as the District of Columbia, Puerto Rico, and the U.S. Virgin
Islands. PCS defines "catastrophes" as events that, in their
estimation, affect a significant number of policyholders and that cause
more than $25 million in damages. To identify catastrophes, PCS reviews
daily weather reports and wire service news stories to determine if
potentially damaging weather has occurred anywhere in the nation. PCS
contacts adjusters, insurance claims departments, or public officials
to gather additional information about the scope of damage and
potential insured losses for events. Damages associated with a single
storm event are grouped together as a single catastrophe, even if they
are separated by distance. PCS obtains its insured loss data from
information reported by insurers. PCS estimates include losses under
personal and commercial property insurance policies covering real
property, contents, business interruption, vehicles, and boats. PCS
estimates also typically include amounts paid by state wind pools,
joint underwriting associations, and certain other residual market
mechanisms, such as Fair Access to Insurance Requirements (FAIR) plans.
However, PCS estimates do not include damage to uninsured or self-
insured property including uninsured publicly owned property and
utilities; losses involving agriculture, aircraft and property insured
under NFIP or certain specialty lines (such as ocean marine), or loss
adjustment expenses. Generally, PCS finalizes its estimates within 6
months of the occurrence of a PCS-identified catastrophe, according to
company documents. PCS does not independently verify or audit the
accuracy of the reported losses. Thus, loss totals are the best
estimates of primary insurers compiled by PCS professionals, and may or
may not accurately and completely reflect actual industry-insured
losses. Nevertheless, PCS has determined their data to be very close to
other independent estimates. PCS officials said that, when compared
with state insurance commissioners' estimates based on all loss data
from insurance companies following particularly large catastrophes, PCS
data are within 3 to 5 percent of actual amounts. For the data used in
our review, company officials told us that most estimates included in
the data provided to us are final, except the 2005 hurricanes.
NFIP:
NFIP data are actual claim payment totals, not estimated amounts. NFIP
data represent the budget outlays that satisfy claims submitted by NFIP
policyholders to their participating program companies. The companies
report these data to the NFIP on a monthly basis. According to a senior
program official, the Department of Homeland Security performs periodic
audits of company records reported to NFIP. Although nearly all claims
in the NFIP data we reviewed are considered closed by the agency (and,
therefore, final), a small portion of claims associated with 2004 and
2005 hurricane season are not reflected in data we reviewed, according
to the agency's database manager.
FCIC:
The loss data provided by FCIC represent the actual amount paid to
policyholders, not estimates. FCIC data represent the budget outlays
that satisfy claims submitted by policyholders to their participating
insurance companies. Participating insurance companies submit claims
information for processing through a computerized validation system.
Automated processing of claims information occurs annually for a period
going back 5 years, but agency officials said that indemnities may have
changed after automated processing closed in very specific cases, such
as settlement of litigation or arbitration cases.
Identifying Insured Losses Associated with Hurricanes:
To determine the insured losses associated with major and nonmajor
hurricanes, we identified losses associated with hurricanes in both the
PCS and NFIP data sets. We used the name and year of each hurricane to
link loss records to information from the National Oceanic and
Atmospheric Administration (NOAA) on the peak intensity of each
hurricane at or near landfall.
Independent Studies:
We supplemented our descriptive analysis with a review of existing
literature and the views of subject area experts on the primary drivers
of changes in the weather-related loss record in general. Given the
data challenges faced by natural hazard researchers, the data sets used
in these studies are generally different.
Interviews with Major Insurers:
To address the third objective, we conducted semistructured interviews
with officials from the NFIP, RMA, and a nonprobability sample of the
largest private property/casualty primary insurance and reinsurance
companies as defined by national market share. In the private sector,
11 out of 14 potential respondents elected to participate, drawing from
companies in the United States, Europe, and Bermuda. Although the
results from this sample should not be generalized to represent all
insurance companies, the companies we interviewed represent about 45
percent of the total domestic insurance market. In developing our
semistructured questionnaire, we reviewed existing literature on risk
assessment and management practices, GAO guidance on risk management,
and interviewed subject area experts knowledgeable about the insurance
industry and federal insurance programs. Insurance industry experts
included representatives from insurance brokers, catastrophe modeling
firms, industry associations, the Insurance Information Institute, and
academics. To reduce response error, we pretested our questions for
clarity, relevancy, and sensitivity with representatives from several
insurance industry associations, including the American Insurance
Association, the National Association of Mutual Insurance Companies,
the Property Casualty Insurance Association of America, and the
Reinsurance Association of America. On the basis of feedback from the
pretests, we modified the questions as appropriate. We distinguished
proactive risk management responses to climate change from other
responses according to whether insurers indicated that they were
adjusting their activities based on projected changes in underlying
weather trends rather than adapting only as changes in weather
conditions reveal themselves in historical data. During our interviews,
some private insurers attributed their actions to changes in the
Atlantic Multidecadal Oscillation (AMO). Because NOAA considers the AMO
to be a climatic cycle, we categorized the actions of these insurers as
responding to climate change.
We asked the participating federal agencies and private insurance and
reinsurance companies to identify individuals knowledgeable about their
weather-related risk management practices for our interviews. Based on
these criteria, we spoke with a range of senior officials and
representatives that included actuaries, underwriters, catastrophe
specialists, regulatory affairs and counsel. During the interviews, we
asked a series of questions about risk assessment and management
practices for weather-related risk, significant drivers of changes to
past and future weather-related risk, respondents' perception of and
actions to address climate change in their risk management processes,
and risk management best practices that might be transferable to
federal insurers.
We also interviewed officials from rating agencies, catastrophe
modeling firms, insurance industry associations, the National
Association of Insurance Commissioners, and universities to provide
additional context for respondents' statements. To supplement our
interviews, we reviewed documentary evidence of risk management
practices from federal agencies, studies from subject area experts,
industry reports, publicly available insurance company documents, and
previous work from GAO to provide context and support for respondents'
statements.
We performed our work between February 2006 and January 2007 in
accordance with generally accepted government auditing standards.
[End of section]
Appendix II: National Flood Insurance Program:
Floods are the most common and destructive natural disaster in the
United States. According to NFIP statistics, 90 percent of all natural
disasters in the United States involve flooding. Because of the
catastrophic nature of flooding and the inability to adequately predict
flood risks, private insurance companies largely have been unwilling to
underwrite and bear the risk of flood insurance. As a result, flooding
is generally excluded from homeowner policies that cover damages from
other types of losses, such as wind, fire, and theft.
The NFIP was established in 1968 to address uninsured losses due to
floods. Prior to the establishment of the NFIP, structural flood
controls on rivers and shorelines (e.g., dams and levees) and disaster
assistance for flood victims were the federal government's primary
tools for addressing floods. The Mississippi River Commission, created
in 1879 to oversee the development of a levee system to control the
river's flow, was the first of these federal efforts to address
flooding. Due to the limited effectiveness of structural flood
controls, continued development in flood-prone areas, and a desire to
reduce postdisaster assistance payments, the Congress began examining
the feasibility of prefunding flood disaster costs via federal
insurance in the 1950s. Although the first federal flood insurance
program authorized by the Congress in 1956 failed due to lack of
funding, a series of powerful hurricanes and heavy flooding on the
Mississippi River in the early 1960s prompted the Congress to revisit
the issue and direct the Department of Housing and Urban Development
(HUD) to conduct a feasibility study of a federal flood insurance
program. The 1966 HUD feasibility study helped lead to the passage of
the National Flood Insurance Act of 1968,[Footnote 48] which authorized
the creation of the NFIP.[Footnote 49]
Since its inception, the NFIP has undergone several major changes in
response to significant flood events. Hurricane Agnes in 1972 led to
the mandatory flood insurance requirements on certain persons in flood-
prone areas included in the Flood Disaster Protection Act of 1973,
which also significantly increased coverage limits in a further effort
to increase participation.[Footnote 50] Following the Midwest floods of
1993, the Congress enacted the National Flood Insurance Reform Act of
1994, which strengthened lender compliance requirements with mandatory
purchase provisions requiring mortgage-holders in flood-prone areas to
purchase flood insurance and prohibited flood disaster assistance for
properties that had not maintained their mandatory coverage.[Footnote
51] In 2004, recognizing that losses from repetitive flooding on some
insured properties was straining the financial condition of the NFIP,
the Congress passed the Flood Insurance Reform Act of 2004, which
provided NFIP with additional tools to reduce the number and financial
impact of these properties.[Footnote 52] These tools include: increased
authorization of funding for mitigation of repetitive loss properties
and statutory authority to penalize policyholders who refuse government
assistance to mitigate certain structures that have been substantially
or repetitively damaged by flooding, among others. Recently, the
Congress has begun exploring additional changes to the NFIP to address
the financial and operational challenges presented by the 2005
hurricane season.
How the Program Works:
FEMA, within the Department of Homeland Security (DHS), is responsible
for the oversight and management of the NFIP.[Footnote 53] Under this
program, the federal government assumes the liability for covered
losses and sets rates and coverage limitations, among other
responsibilities.
The NFIP combines three elements: (1) property insurance for potential
flood victims, (2) mapping to identify the boundaries of the areas at
highest risk of flooding, and (3) incentives for communities to adopt
and enforce floodplain management regulations and building standards
(such as elevating structures) to reduce future flood damage. The
effective integration of all three of these elements is needed for the
NFIP to achieve its goals of:
* providing property flood insurance coverage for a high proportion of
property owners who would benefit from such coverage,
* reducing taxpayer-funded disaster assistance when flooding strikes,
and:
* reducing flood damage through floodplain management and the
enforcement of building standards.
Over 20,000 communities across the United States and its territories
participate in the NFIP by adopting and agreeing to enforce state and
community floodplain management regulations to reduce future flood
damage. In exchange, the NFIP makes federally backed flood insurance
available to homeowners and other property owners in these communities.
As of 2005, the program had over 4.9 million policyholders,
representing about $875 billion in assets. Homeowners with mortgages
from federally regulated lenders on property in communities identified
to be in high flood risk areas are required to purchase flood insurance
on their dwellings. Optional, lower cost coverage is also available
under the NFIP to protect homes in areas of low to moderate risk. The
mandated coverage protects homeowners' dwellings only; to insure
furniture and other personal property items against flood damage,
homeowners must purchase separate NFIP personal property coverage.
Prior to the 2005 hurricanes, NFIP had paid about $14.6 billion in
flood insurance claims, primarily from policyholder premiums that
otherwise would have been paid through taxpayer-funded disaster relief
or borne by home and business owners themselves. According to FEMA,
every $3 in flood insurance claims payments saves about $1 in disaster
assistance payments, and the combination of floodplain management and
mitigation efforts save about $1 billion in flood damage each year.
To make flood insurance available on "reasonable terms and conditions
to persons who have need for such protection,"[Footnote 54] the NFIP
strikes a balance between the scope of the coverage provided and the
premium amounts required to provide that coverage. Policy coverage
limits arise from statute and regulation, including FEMA's standard
flood insurance policy (SFIP), which is incorporated in regulation and
issued to policyholders when they purchase flood insurance. As of 2006,
FEMA estimated 26 percent of its policies were subsidized, and 74
percent were charged "full-risk premium" rates. In 1981, FEMA set the
operating goal of generating premiums at least sufficient to cover
losses and expenses relative to the "historical average loss year."
However, the heavy losses from the 2005 hurricane season may increase
the historical average loss year to a level beyond the expected long-
term average. In light of this, FEMA is currently revisiting the use of
the historical average loss year as a premium income target.
Risk Assessment Practices:
The NFIP uses hydrologic models to estimate loss exposure in flood-
prone areas, based on the method outlined in the 1966 HUD report,
Insurance and Other Programs for Financial Assistance to Flood
Victims.[Footnote 55] These techniques of analysis were first developed
by hydrologists and hydraulic engineers to determine the feasibility of
flood protection.
The hydrologic method uses available data on the occurrence of floods
and flood damages to establish both the frequency of flood recurrence
and the damage associated with a flood of a given height. The NFIP
augments available flood data with detailed engineering studies,
simulations, and professional judgment to establish the scientific and
actuarial basis for its risk assessment process and rates.
Flood-elevation frequency data for specific communities is published in
Flood Rate Insurance Maps, which differentiate areas based on their
flood risk. These maps are the basis for setting insurance rates,
establishing floodplain management ordinances, and identifying
properties where flood insurance is mandatory.
To estimate expected annual losses and determine the basis for rate
setting, NFIP combines flood-elevation frequency data with depth-damage
calculations to estimate a range of flood probabilities and associated
damages. Each possible flood is multiplied by the expected damage
should such a flood occur, and then each of these is added together.
The total of each possible flood's damage provides an expected per
annum percentage of the value of property damage due to flooding. This
expected damage can then be converted to an expected loss per $100 of
property value covered by insurance. This per annum expected loss
provides the fundamental component of rate setting. Rates are also
adjusted to incorporate additional expense factors, such as adjustment
costs and deductibles.
Program Funding:
To the extent possible within the context of its broader purposes, the
NFIP is expected to pay operating expenses and flood insurance claims
with premiums collected on flood insurance policies rather than with
tax dollars. However, as we have reported, the program is not
actuarially sound by design because the Congress authorized subsidized
insurance rates to be made available for policies covering certain
structures to encourage communities to join the program. As a result,
the program does not collect sufficient premium income to build
reserves to meet the long-term future expected flood losses.[Footnote
56] FEMA has statutory authority to borrow funds from the Department of
the Treasury to keep the NFIP solvent.[Footnote 57] Prior to the 2005
hurricane season, FEMA had exercised its borrowing authority four
times, when losses exceeded available fund balances. For example, FEMA
borrowed $300 million to pay an estimated $1.8 billion on flood
insurance claims resulting from the 2004 hurricane season. Following
hurricanes Katrina, Rita, and Wilma, FEMA estimates it will need to
borrow nearly $21 billion dollars to cover outstanding claims. Although
FEMA has repaid borrowed funds with interest in the past, FEMA does not
expect to be able to meet the $1 billion in annual interest payments
for these borrowed funds.
[End of section]
Appendix III: Federal Crop Insurance Corporation:
In general, farm income is determined on the basis of farm production
and prices, both of which are subject to wide fluctuations due to
external factors. Because a substantial part of farming depends on
weather, farm production levels can vary substantially on an annual
basis. Commodity prices are also subject to significant swings due to
supply and demand on the domestic and international markets. The
Congress created FCIC in 1938 to administer a federal crop insurance
program on an experimental basis to temper the weather effects of the
dust bowl and the economic effects of the Great Depression.[Footnote
58]
The federal crop insurance program protects participating farmers
against financial losses caused by droughts, floods, or other natural
disasters. Until 1980, the federal crop insurance program was limited
to major crops in the nation's primary production areas. The Federal
Crop Insurance Act of 1980 expanded crop insurance both in terms of
crops and geographic areas covered.[Footnote 59] The expansion was
designed to allow the disaster assistance payment program provided by
the government under previous farm bills to be phased out. To encourage
participation, the 1980 act required a 30 percent premium subsidy for
producers who purchased coverage up to the 65 percent yield level.
Despite the subsidies, program participation remained low, and the
Congress authorized several ad hoc disaster payments between 1988 and
1993. Congressional dissatisfaction with the size and frequency of
these payments prompted the Congress to pass the Federal Crop Insurance
Reform Act of 1994, which mandated participation in the crop insurance
program as a prerequisite for other benefits, including agriculture
price support payments.[Footnote 60] The 1994 act also introduced
catastrophic risk protection coverage, which compensated farmers for
losses exceeding 50 percent of their average yield at 60 percent of the
commodity price. Premiums for catastrophic risk protection coverage
were completely subsidized, and subsidies for other coverage levels
were also increased.
As part of the 1996 Farm Bill, the Congress created the Office of Risk
Management under the U.S. Department of Agriculture (USDA), and USDA
established RMA to administer the FCIC insurance programs, among other
things.[Footnote 61] The Congress also required the creation of a
revenue insurance pilot project and repealed the mandatory
participation provision of the 1994 Act. However, participation in the
crop insurance program has not necessarily precluded the need for
further disaster assistance. For example, due to low commodity prices
in 1997 and multiple years of natural disasters, the Congress enacted
an emergency farm financial assistance package totaling almost $6
billion in 1998, which included over $2 billion in crop disaster
payments, and an $8.7 billion financial assistance package in 1999 that
included $1.2 billion in crop disaster payments.
In 2000, the Congress enacted the Agricultural Risk Protection Act,
which further increased subsidies for insurance above the catastrophic
risk protection coverage level, subsidized a portion of the cost of
revenue insurance products, improved coverage for farmers affected by
multiple years of natural disasters, required pilot insurance programs
for livestock farmers, and authorized pilot programs for growers of
other commodities not currently covered, gave the private sector
greater representation on the FCIC Board of Directors, reduced
eligibility requirements for permanent disaster payment programs for
noninsured farmers, and provided new tools for monitoring and
controlling program abuses, among other provisions.[Footnote 62] These
changes required $8.2 billion in additional spending from fiscal years
2001 through 2005.
How the Program Works:
RMA has overall responsibility for supervising the federal crop
insurance program, which it administers in partnership with private
insurance companies. Insurance policies are sold and completely
serviced through approved private insurance companies that have their
losses reinsured by USDA. These companies share a percentage of the
risk of loss or opportunity for gain associated with each insurance
policy written. In addition, RMA pays companies a percentage of the
premium on policies sold to cover the administrative costs of selling
and servicing these policies. In turn, insurance companies use this
money to pay commissions to their agents who sell the policies and fees
to adjusters when claims are filed. RMA oversees the development of new
insurance products and the expansion of existing insurance products to
new areas to help farmers reduce the chance of financial loss.
The USDA determines whether the federal crop insurance program will
insure a commodity on a crop-by-crop and county-by-county basis, based
on farmer demand for coverage and the level of risk associated with the
crop in the region, among other factors. Over 100 crops are covered;
major crops such as grains are covered in almost every county where
they are grown, and specialty crops such as fruit are covered in some
areas. For many commodities, producers may also purchase revenue
insurance. Based on commodity market prices and the producer's
production history, producers are assigned a target revenue level. The
producer receives a payment if their actual revenue falls short of the
target level, whether the shortfall was due to low yield or low prices.
Premiums for revenue insurance are subsidized at the same level as
traditional crop insurance policies.
Farmers' participation in the federal crop insurance program is
voluntary, but the federal government encourages it by subsidizing the
insurance premiums. Participating farmers are assigned a "normal" crop
yield based on their past production history and a commodity price
based on estimated market conditions. The producer selects both the
percentage of yield to be covered and the percentage of the commodity
price received as payment if the producer's losses exceed the selected
threshold. Premium prices increase as levels of yield and price
coverage rise. However, all eligible producers can receive fully
subsidized catastrophic risk protection coverage that pays producers
for losses exceeding 50 percent of normal yield, at a level equal to 55
percent of the estimated market price, in exchange for a $100
administrative fee. Producers who purchase this coverage can buy
additional insurance at partially subsidized rates up to 85 percent of
their yield and 100 percent of the estimated market price.
As an alternative, the Group Risk Plan provides coverage based on
county yields rather than a producer's actual production history. If
county yield falls below the producer's threshold yield (a percentage
of the historical county yield), then the producer receives a payment.
Risk Assessment Practices:
RMA's risk assessment/rate-setting methodology is complex because the
risk of growing a particular crop varies by county, farm, and farmer.
Because of all the possible combinations involved, hundreds of
thousands of rates are in place. Each year, RMA follows a multistep
process to establish rates for each crop included in the program. The
process involves establishing base rates for each county crop
combination and adjusting these basic rates for a number of factors,
such as coverage and production levels. In addition, rates are adjusted
to account for the legislated limitations in price increases.
For each crop, RMA extracts data on counties' crop experience from its
historical database. The data elements for each crop, crop year, and
county include (1) the dollar amount of the insurance coverage sold,
(2) the dollar amount of the claims paid, and (3) the average coverage
level. The historical data are adjusted to the 65 percent coverage
level (the most commonly purchased level of coverage) so that liability
and claims data at different coverage levels can be combined to develop
rates. Using the adjusted data, FCIC computes the loss-cost ratio for
each crop in each county. The loss-cost ratio is calculated by dividing
the total claim payments by the total insurance in force; the result is
stated as a percentage.[Footnote 63] To reduce the impact a single year
will have on the average loss-cost ratio of each county, RMA caps the
adjusted average loss-cost ratio for any single year at 80 percent of
all years.[Footnote 64] To establish the base rate for each county, the
average for all the years since 1975 is calculated using the capped
loss-cost ratios and a weighting process to minimize the differences in
rates among counties.
Rates are further adjusted by: a disaster reserve factor, a surcharge
for catastrophic coverage for each crop based on pooled losses at the
state level,[Footnote 65] a prevented planting factor, farm divisions,
crop type, and differences in both average yield and coverage
levels.[Footnote 66]
Program Funding:
The crop insurance program is financed primarily through general fund
appropriations and farmer-paid premiums. In addition to the premiums
paid by producers, FCIC receives an annual appropriation to cover
necessary costs for the program's premium subsidies, excess losses,
delivery expenses, and other authorized expenses. According to USDA
budget documents, for fiscal year 2005, insurance premium and
administrative fee revenue from farmers was approximately $2.1 billion,
and gross claims equaled almost $3.3 billion. Total government
operating costs in fiscal year 2005 were approximately $3 billion.
RMA is required to set crop insurance premiums at actuarially
sufficient rates, defined as a long-run loss ratio target of no more
than 1.075. From its initial expansion in 1981 through 1994, the crop
insurance program had an average loss ratio of 1.47 and paid roughly
$3.2 billion in claims excess of subsidized premium income during that
period.[Footnote 67] From 1995 to 2005, the program had an average loss
ratio of 0.91, and collected roughly $2.7 billion in subsidized premium
excess of claims during that period. Excluding subsidies and measuring
performance on the basis of a producer premium, from 1981 to 1994, the
crop insurance program averaged a loss ratio of 1.93 and paid roughly
$5.2 billion in claims excess of producer premium over that period;
from 1995 to 2005, the program averaged a loss ratio of 2.15 and paid
roughly $14.2 billion in claims excess of a producer premium during
that period.
Generally, producers can purchase crop insurance to insure up to 85
percent of their normal harvest (yield), based on production history.
In 2007, the USDA expects the FCIC to provide $48 billion in risk
protection on 287 million acres nationwide, which represents
approximately 80 percent of the nation's acres planted to principal
crops. The USDA estimates this level of coverage will cost the federal
government $4.2 billion in 2007.
[End of section]
Appendix IV: Consensus Statement among Participants at 2006 Munich Re
Workshop:
Munich Re, one of the world's largest reinsurance companies, and the
University of Colorado jointly convened an international workshop on
climate change and disaster loss trends in May 2006 in Hohenkammer,
Germany. The workshop brought together 32 experts in the fields of
climatology and disaster research from 13 countries. White papers were
prepared and circulated by 25 participants in advance of the workshop
and formed the basis of the discussions. In the course of the event,
participants developed a list of statements that each represent a
consensus among participants on issues of research and policy as
related to the workshop's two central organizing questions: (1) What
factors account for increasing costs of weather related disasters in
recent decades? and (2) What are the implications of these
understandings, for both research and policy?
Consensus (unanimous) statements of the workshop participants:
1. Climate change is real, and has a significant human component
related to greenhouse gases.
2. Direct economic losses of global disasters have increased in recent
decades with particularly large increases since the 1980s.
3. The increases in disaster losses primarily result from weather
related events, in particular storms and floods.
4. Climate change and variability are factors which influence trends in
disasters.
5. Although there are peer reviewed papers indicating trends in storms
and floods there is still scientific debate over the attribution to
anthropogenic climate change or natural climate variability. There is
also concern over geophysical data quality.
6. IPCC (2001) did not achieve detection and attribution of trends in
extreme events at the global level.
7. High quality long-term disaster loss records exist, some of which
are suitable for research purposes, such as to identify the effects of
climate and/or climate change on the loss records.
8. Analyses of long-term records of disaster losses indicate that
societal change and economic development are the principal factors
responsible for the documented increasing losses to date.
9. The vulnerability of communities to natural disasters is determined
by their economic development and other social characteristics.
10. There is evidence that changing patterns of extreme events are
drivers for recent increases in global losses.
11. Because of issues related to data quality, the stochastic nature of
extreme event impacts, length of time series, and various societal
factors present in the disaster loss record, it is still not possible
to determine the portion of the increase in damages that might be
attributed to climate change due to greenhouse gas emissions.
12. For future decades the IPCC (2001) expects increases in the
occurrence and/or intensity of some extreme events as a result of
anthropogenic climate change. Such increases will further increase
losses in the absence of disaster reduction measures.
13. In the near future the quantitative link (attribution) of trends in
storm and flood losses to climate changes related to greenhouse gas
emissions is unlikely to be answered unequivocally.
14. Adaptation to extreme weather events should play a central role in
reducing societal vulnerabilities to climate and climate change.
15. Mitigation of greenhouse gas emissions should also play a central
role in response to anthropogenic climate change, though it does not
have an effect for several decades on the hazard risk.
16. We recommend further research on different combinations of
adaptation and mitigation policies.
17. We recommend the creation of an open-source disaster database
according to agreed upon standards.
18. In addition to fundamental research on climate, research priorities
should consider needs of decision makers in areas related to both
adaptation and mitigation.
19. For improved understanding of loss trends, there is a need to
continue to collect and improve long-term and homogenous data sets
related to both climate parameters and disaster losses.
20. The community needs to agree upon peer reviewed procedures for
normalizing economic loss data.
[End of section]
Appendix V: Comments from the U.S. Department of Agriculture:
Note: GAO comments supplementing those in the report text appear at the
end of this appendix.
USDA:
United States Department of Agriculture:
Office of the Secretary:
Washington, D.C. 20250:
Feb 23 2007:
Mr. John B. Stephenson:
Director:
Natural Resources and Environment:
Government Accountability Office:
441 G Street N.W.
Washington, D.C. 20548:
Dear Mr. Stephenson:
Enclosed is the Farm and Foreign Agricultural Service's response to the
draft report titled, Climate Change: Financial Risks to Federal and
Private Insurers in Coming Decades are Potentially Significant." Thank
you for the opportunity to provide comments. If you have any questions
regarding our response, please contact Michael Hand at 202-720-0642.
Sincerely,
Signed by:
Mark Keenum:
Under Secretary:
Farm and Foreign Agricultural Services:
Enclosure:
U.S. Department of Agriculture Response to the U.S. Government
Accountability Office Draft Report GAO-07-285 "Climate Change:
Financial Risks to Federal and Private Insurers in Coming Decades are
Potentially Significant"
February 8, 2007:
Weather-related events have caused billions of dollars in damage over
the past decade. GAO examined actions taken by private and Federal
insurers to address the potential increase in losses. As a result of
the study, GAO recommends that the United States Department of
Agriculture (USDA), specifically the Risk Management Agency (RMA),
analyze the potential long-term implications of climate change using
forthcoming assessments from the Intergovernmental Panel on Climate
Change to establish sound estimates of expected future conditions.
USDA Response:
USDA is in general agreement with GAO's recommendation.
Specific Comments:
Although USDA agrees with GAO's recommendation, we do not agree with
some of the conclusions drawn within the report.
Much of the focus of this report is with losses related to coastal
weather events, especially hurricanes. However, the main cause of
catastrophic losses for the crop insurance program is drought in the
nation's interior. This is why the loss experience of the crop
insurance program is distinct from the loss experience described in the
report for the National Flood Insurance Program and property and
casualty losses for private insurers.
The increase in crop insurance indemnities over time reflects the rapid
growth of the crop insurance program, not an increase in either the
frequency and/or severity of catastrophic weather events. In fact, the
severity of loss for the crop insurance program, as measured by the
loss ratio, has been generally lower in the 1990's and 2000's than in
the 1980's. Thus, if anything, the frequency and severity of
catastrophic loss events for the crop insurance program appears to be
decreasing.
USDA does not agree that it has "taken little action to prospectively
assess potential increases in catastrophic risk associated with climate
change." RMA tracks total program liability - a definitive measure of
the total value at risk from climatic weather events. This number is
updated on a weekly basis and is available on RMA's public website.
RMA also estimates expected changes in liability up to 10 years ahead
through RMA's baseline projections. Therefore, RMA does assess the long-
term, as well as current, exposure of the crop insurance program to
catastrophic weather events.
GAO's draft report treats the recurring 20-to 40-year Atlantic
hurricane cycle as synonymous with climate change. However, other parts
of the report describe climate change in terms of a long-run
progression, such as global warming, that leads to an increase in
frequency and severity of weather events. Referring to the normal cycle
of Atlantic hurricanes as climate change appears to be inconsistent
with how climate change is described in other parts of the report.
When GAO surveyed private insurers about what they are doing to
estimate and prepare for the risks of climate change, they found that
insurers were using catastrophe models that incorporate the hurricane
cycle. RMA also incorporates hurricane risk into premium rates for
several of its insured commodities. However, rather than focusing on
short-term fluctuations in the hurricane cycle, RMA uses historical
hurricane data that spans several cycles.
The following are GAO's comments on the U.S. Department of
Agriculture's letter dated February 23, 2007.
GAO Comments:
1. We agree that the loss experiences of NFIP, FCIC, and private
insurers are distinct and sought to reflect these distinctions in our
draft report. For example, we acknowledged on page 23 of the draft the
specific distinction USDA highlights--that the main cause of
catastrophic losses for FCIC is drought in the nation's interior (see
pages 24 and 25 of this document). Despite these and other differences,
however, we believe the report's findings and underlying message are
still applicable to the NFIP, the FCIC, and private insurers.
2. Our analysis of insured losses does not attempt to attribute
increases in past losses to changes in the severity of weather events
in the data sets we reviewed, as implied by the comment. Moreover, we
acknowledge that the increase in FCIC's losses (indemnities) largely
reflected the rapid growth of the crop insurance program. However,
given the IPCC's projections for potential increase in the frequency
and severity of weather-related events--including those that affect
crops--we believe that limiting an evaluation of FCIC's future weather-
related risk to the program's loss ratio--which only captures
historical performance of the program based on past climatic and market
conditions--to be a potentially misleading metric upon which to make a
prospective assessment.
3. We acknowledged these activities in the draft report. However, we
believe that USDA's actions are limited in scope, focusing almost
exclusively on actuarial performance and not on the potential
implications of climate change for FCIC's operations (i.e., changes in
the frequency and severity of weather-related events, weather
variability, growing seasons, and pest infestations). Accordingly, we
believe the program should do more to prospectively assess the
implications of climate change.
4. We employed the IPCC's definition of climate change, which includes
statistically significant variations in climate, brought on by factors
that are both internal and external to the earth's climate system, and
that persist over time--typically decades or longer. Under this
definition, the Atlantic hurricane cycle, as with other significant
variations that are understood to be internal to the earth's climate
system, can be considered climatic changes. Our use of the definition
was corroborated by a senior NOAA scientist.
5. We updated our discussion of FCIC's modeling activities (see page
36) to reflect this hurricane model. However, as stated on page 22, 75
percent of FCIC's claims were associated with drought, excess moisture,
and hail from 1980 to 2005, whereas hurricanes were associated with a
much smaller portion of FCIC's claims during this period. Accordingly,
we believe that if more sophisticated, prospective risk assessment
techniques (such as those used in FCIC's hurricane model) were applied
to drought, moisture, and hail events, it would allow for a far more
useful assessment of the potential implications of climate change for
FCIC's operations.
[End of section]
Appendix VI: Comments from the Department of Commerce:
Note: GAO comments supplementing those in the report text appear at the
end of this appendix.
United States Department Of Commerce:
The Under Secretary of Commerce for Oceans and Atmosphere:
Washington, D.C. 20200:
Feb 26 2007:
Mr. John B. Stephenson:
Director:
Natural Resources and Environment:
U.S. Government Accountability Office:
441 G Street, N W:
Washington, D.C. 20548:
Dear Mr. Stephenson:
Thank you for the opportunity to review and comment on the Government
Accountability Office's draft report entitled Climate Change: Financial
Risks to Federal and Private Insurers In Coming Decades arc Potentially
Significant (GAO-07-285). Enclosed is the National Oceanic and
Atmospheric Administration's comments on the draft report.
Sincerely,
Signed by:
Conrad C. Lautenbacher, Jr.
Vice Admiral, U.S. Navy (Ret.):
Under Secretary of Commerce for Oceans and Atmosphere:
Enclosure:
Department of Commerce National Oceanic and Atmospheric Administration
Comments on the Draft GAO Report Entitled "Climate Change: Financial
Risks to Federal and Private Insurers In Coming Decades are Potentially
Significant" (GAO-07-285/March 2007):
General Comments:
The Department of Commerce (DOC) appreciates the opportunity to review
this report. The issues covered in the report are very important and
reflect the real world intersection between science, policy, and
economics.
We have three major comments on the structure of the report. First, GAO
should provide a clear definition of the phrase "climate change" at the
beginning of its report. While it is addressed on page 2, DOC
recommends the authors refer to the definition provided by the 2007
Intergovernmental Panel on Climate Change (IPCC) Working Group 1:
IPCC Working Group 1 Climate Change Definition:
Climate change refers to a change in the state of the climate that can
be identified (e.g., using statistical tests) by changes in the mean
and/or the variability of its properties, and that persists for an
extended period, typically decades or longer. Climate change may be due
to natural internal processes or external forcings, or to persistent
anthropogenic changes in the composition of the atmosphere or in land
use.
The second comment is directed to page 2 of the report and relates to
the discussion of frequency and intensity of weather phenomenon. The
authors write:
"Regardless of the cause, increasing temperatures accompanied by
changes in other aspects of the climate-may impact communities and, by
extension, the insurance industry by altering the frequency or severity
of weather-related events such as hurricanes, tornadoes, severe
thunderstorms and hail events, and wildfires."
While DOC recognizes the IPCC's Fourth Assessment Report was not
available at the time of GAO's review, the issue of frequency and
intensity has been well discussed in the scientific community, and
policy makers would benefit from drawing information from the IPCC's
Summary for Policy Makers for Working Group 1. According to page 10 of
this summary, "there is insufficient evidence to determine whether
trends exist . in small scale phenomena such as tornadoes, hail,
lightning, and dust storms." The authors could state the frequency of
heavy precipitation events has increased over most land areas. (page
8). On hurricanes, IPCC notes an increase in "intense tropical cyclone
activity," but also mentions "there is no clear trend in the annual
numbers of tropical cyclones," which refers to frequency. Tropical
cyclones projections are addressed on page 16 of the summary, where the
IPCC projects future tropical cyclones will become more intense, but
there is less confidence in projections of a global decrease in numbers
of tropical cyclones.
Further, DOC notes the report could be strengthened by a discussion or
what is meant by "altering the frequency or severity of weather-related
events" and how this is linked to risk. For example, altering either
the frequency or severity of high impact extreme weather-related events
could result in a five fold increase in risk for what has been
considered a 500-year event (i.e., probability of occurring in a given
year = 1/500) shifts under climate change to be a 100-year event (i.e.,
probability of occurring in a given year = 1/100).
The third comment is the report should examine coastal development
impacts more rigorously. The National Oceanic and Atmospheric
Administration (NOAA) has done work that uses data from the Bureau of
the Census to show coastal communities have seen population growth of
nearly 40 million people from 1970 to 2000. The authors refer to Roger
Pielke, Jr.'s work on coastal impacts, but cite it only to show that
more intense hurricanes tend to have higher impacts. Pielke, Jr., and
others, including Chris Landsea of NOAA and Kerry Emanuel of
Massachusetts Institute of Technology, have examined hurricanes,
climate change, and development, and found coastal development has
increased the vulnerability to winter storm surge, wind damage, and
hurricanes. "These vulnerabilities, due to high risk coastal
development, will only be amplified by climate change related increases
in the frequency or severity of high impact extreme weather-related
events.
The authors cite anecdotal evidence, such as increased development in
the area hit by Hurricane Andrew, but the report lacks analysis of the
long term trends and does not quantify what portion of the increase in
losses is attributable to societal change and economic development as
referenced on page 58 in the Munich Re consensus statement. This would
be useful information for policy makers.
The following are GAO's comments on the Department of Commerce's letter
dated February 26, 2007.
GAO Comments:
1. We agree that a clear and accurate definition of climate change is a
necessary prerequisite for any discussion of the issue. While a variety
of definitions for the term are in use, we did not attempt to
independently define the term. Rather, we relied upon the IPCC's most
current publicly-available definition.
2. We revised the introductory statement referred to in Commerce's
comments for editorial purposes (see page 2). To the extent
practicable, we also incorporated the Working Group I Summary for
Policymakers of the IPCC's Fourth Assessment Report into the detailed
discussion of the potential changes in the frequency and severity of
weather-related events identified in the 2001 Third Assessment Report
(see pages 8 to 13).
3. We included an elaboration on page 14 of how altering the frequency
and severity of weather-related events is linked to risk.
4. It was outside the scope of this report to conduct our own
quantitative trend analysis of the relative roles of societal factors
(such as development or agricultural prices) and climate change in
shaping the increases in weather-related insured losses observed in the
data. In response to the comment, however, we clarified which studies
we reviewed that addressed this question, both for coastal hazards
(such as hurricanes) and inland hazards (such as drought and excess
moisture).
[End of section]
Appendix VII: GAO Contact and Staff Acknowledgments:
GAO Contact:
John Stephenson, (202) 512-3841, or stephensonj@gao.gov:
Staff Acknowledgments:
In addition to the individual named above, Steve Elstein, Assistant
Director; Chase Huntley; Alison O'Neill; Michael Sagalow; and Lisa Van
Arsdale made key contributions to this report. Charles Bausell, Jr;
Christine Bonham; Mark Braza; Lawrence Cluff; Arthur James, Jr; Marisa
London; Justin Monroe; and Greg Marchand also made important
contributions to this report.
We also wish to give special tribute to our dear friend and colleague,
Curtis Groves, who died many years too soon after a long battle with
multiple myeloma near the conclusion of our work.
[End of section]
Related GAO Products:
National Flood Insurance Program: New Processes Aided Hurricane Katrina
Claims Handling, but FEMA's Oversight Should Be Improved. GAO-07-169.
Washington, D.C.: December 15, 2006.
Catastrophic Disasters: Enhanced Leadership, Capabilities, and
Accountability Controls Will Improve the Effectiveness of the Nation's
Preparedness, Response, and Recovery System. GAO-06-618. Washington,
D.C.: September 6, 2006.
Crop Insurance: More Needs To Be Done to Reduce Program's Vulnerability
to Fraud, Waste, and Abuse. GAO-06-878T. Washington, D.C.: June 15,
2006.
High-Risk Program. GAO-06-497T. Washington, D.C.: March 15, 2006.
Federal Emergency Management Agency: Improvements Needed to Enhance
Oversight and Management of the National Flood Insurance Program. GAO-
06-119. Washington, D.C.: October 18, 2005.
Crop Insurance: Actions Needed to Reduce Program's Vulnerability to
Fraud, Waste, and Abuse. GAO-05-528. Washington, D.C.: September 30,
2005.
Catastrophe Risk: U.S. and European Approaches to Insure Natural
Catastrophe and Terrorism Risks. GAO-05-199. Washington, D.C.: February
28, 2005.
21st Century Challenges: Reexamining the Base of the Federal
Government. GAO-05-325SP. Washington, D.C.: February 2005.
Climate Change: Information on Three Air Pollutants' Climate Effects
and Emissions Trends. GAO-03-25. Washington, D.C.: April 28, 2003.
FOOTNOTES
[1] Insurers use the term "loss" to refer to the dollar value of
approved or settled claims arising from damages incurred by a
policyholder. For the purposes of this report, weather-related loss
refers to the dollar value of claims made on damage attributable to
weather-related events. "Loss" does not account for premium or other
income, deductibles, co-payments, or damages in excess of coverage.
[2] More specifically, we used the Intergovernmental Panel on Climate
Change definition, which refers to climate change as a statistically
significant variation in either the mean state of the climate or in its
variability, persisting for an extended period (typically decades or
longer). Climate change may be due to natural factors (e.g., internal
processes or external forcings such as solar variations or heavy
volcanic activity), or to persistent human-induced changes in the
composition of the atmosphere or land use patterns.
[3] The National Association of Insurance Commissioners is an
organization of insurance regulators from the 50 states, the District
of Columbia, and the five U.S. territories.
[4] Federal insurance programs are not designed to earn financial
profits.
[5] To insure a risk, private insurers must be able to both estimate an
event's occurrence and its associated damages and be able to set
premiums sufficient to cover their risk and earn a profit. In some
cases, insurers may be prevented from charging sufficient premiums due
to state regulatory actions.
[6] See appendixes II and III for additional information on how these
programs operate, how they assess risk, and how they are funded.
[7] Appendix I contains additional information on the specific
assessments we reviewed. CCSP is a multiagency effort to coordinate
federal climate change science that is responsible for preparing a
series of 21 climate science synthesis and assessment products (SAP)
for the United States by 2008.
[8] This estimate comes from a recently released summary of a key
component of IPCC's Fourth Assessment Report of the state of climate
science, which reported an updated 100-year linear trend (1906-2005) of
0.74 degrees Celsius--larger than the corresponding 0.6 degrees Celsius
reported in the 2001 Third Assessment Report.
[9] For the purposes of this report, extreme weather-related events are
those with a low frequency of occurrence, but that cause severe damage,
such as hurricanes, drought, winter storms, tornadoes, wildfires, and
floods, among others.
[10] IPCC narrowed its range of projected warming in its recently
released summary from the corresponding range of 1.4 to 5.8 degrees
Celsius reported in the 2001 Third Assessment Report. Although these
two sets of projections are broadly consistent, they are not directly
comparable. IPCC notes in the summary that the new range is more
advanced in that it provides best estimates and an assessed likelihood
range. It also relies on a larger number of climate models of
increasing complexity and realism, as well as new information regarding
the nature of feedbacks from the carbon cycle and constraints on
climate response from observations.
[11] National Research Council, Surface Temperature Reconstructions for
the Last 2,000 Years (Washington, D.C.: 2006).
[12] Likelihoods for projected changes are defined by the following
conditions set by the IPCC: "very likely" indicates that a number of
models have been analyzed for such a change, all those analyzed show it
in most regions, and it is physically plausible; and "likely" indicates
that theoretical studies and those models analyzed show such a change,
but only a few models are configured in such a way as to reasonably
represent such changes.
[13] The most recent national assessment for the United States,
entitled Climate Change Impacts on the United States, was forwarded by
a federal advisory committee to the Congress and the President in 2000
as required by the Global Change Research Act of 1990. We reported in
2005 that the subsequent assessment was not submitted in November 2004
as required by the act. Instead, according to the Department of
Commerce, CCSP has committed to issuing 21 shorter reports by 2008. See
GAO, Climate Change Assessment: Administration Did Not Meet Reporting
Deadline, GAO-05-338R (Washington, D.C.: Apr. 14, 2005).
[14] Andrew Dlugolecki, The Changing Risk Landscape: Implications for
Insurance Risk Management (1999) Hyperlink,
http://www.aon.com.au/pdf/reinsurance/Aon_Climate_Change.pdf
(downloaded Jan. 8, 2007).
[15] See Roger Pielke, Jr., et al., Normalized Hurricane Damages in the
United States: 1900-2005 (2007), accessed via Hyperlink,
http://sciencepolicy.colorado.edu/publications/special/normalized_hurric
ane_damages.html (downloaded Jan. 8, 2007).
[16] The Saffir-Simpson hurricane intensity category system was
developed in the 1970s to calculate the destructive force of
hurricanes. The scale ranges from Category One to Category Five, with
Category Five being the most severe. For example, Category Three
hurricanes have winds of 111 to 130 mph, whereas Category Five
hurricanes have winds greater than 155 mph.
[17] Data throughout this section are presented in constant 2005
dollars to allow for a comparison of the dollar value of losses over
time and are not otherwise adjusted. See appendix I for more
information on data used in this report.
[18] Property Claim Services (PCS), an authority on insured property
losses, maintains a database of estimated losses determined to be
"catastrophes"--that is, loss events larger than $25 million that
affect a significant number of policyholders. PCS estimates include
losses under personal and commercial property insurance policies and
typically include payments made on behalf of state-administered risk
pools. PCS data are described in greater detail in appendix I.
[19] Appendixes II and III provide additional information about the
structure and operation of FCIC and NFIP. Importantly, totals only
reflect what was paid during this time--some losses incurred in 2005
may be omitted from this data set.
[20] Weather-related damages are also responsible for many indirect and
non-market impacts that are not entirely accounted for, if at all, in
economic terms, such as environmental damage. See NAS, The Impacts of
Natural Disasters: A Framework for Loss Estimation (Washington, D.C.:
1999), 55-64.
[21] GAO, Catastrophe Risk: U.S. and European Approaches to Insure
Natural Catastrophe and Terrorism Risks, GAO-05-199 (Washington, D.C.:
Feb. 28, 2005), 61.
[22] Estimate was produced by Marshall & Swift/Boeckh, a leading
supplier of local building cost information, residential and commercial
property valuation services for the property and casualty insurance
sector in the United States. GAO did not independently evaluate the
reliability of this estimate.
[23] Munich Re, Topics 2000: Natural Catastrophes--the Current
Position. Geoscience Research Group (Munich, Germany: 1999).
[24] NHC estimates total losses by extrapolating from insured losses by
assuming they account for approximately 50 percent of total losses.
[25] NAS (1999), 1.
[26] A normalization provides an estimate of the damage that would
occur if storms from the past affected the same location under the
societal conditions of another year.
[27] See, for example, Roger A. Pielke, Jr., "Disasters, Death, and
Destruction: Making Sense of Recent Calamities," Oceanography, vol. 19,
no. 2 (2006); Stanley A. Changnon et al., "Human Factors Explain the
Increased Losses from Weather and Climate Extremes," Bulletin of the
American Meteorological Society, vol. 81, no. 3 (2000); and Roger A.
Pielke, Jr., and Christopher W. Landsea, "Normalized Hurricane Damages
in the United States: 1925-95," Weather and Forecasting, vol. 13
(1998).
[28] See, for example, Evan Mills, Richard J. Roth, Jr., and Eugene
Lecomte, Availability and Affordability of Insurance Under Climate
Change: A Growing Challenge for the U.S. (Boston, Mass.: December
2005); Paul Epstein and Evan Mills, eds., Climate Change Futures:
Health, Ecological, and Economic Dimensions (Boston, Mass.: November
2005); and Cynthia Rosenzweig et al., "Increased Crop Damage in the
U.S. from Excess Precipitation Under Climate Change," Global
Environmental Change, vol. 12 (2002).
[29] Peter Höppe and Roger Pielke, Jr., eds., Report of the Workshop on
Climate Change and Disaster Losses: Understanding and Attributing
Trends and Projections, Hohenkammer, Germany, May 25-26, 2006 (Munich,
Germany: October 2006).
[30] Consensus statements agreed to at the workshop are listed in their
entirety in appendix IV.
[31] GAO, Catastrophe Insurance Risks: The Role of Risk-Linked
Securities and Factors Affecting Their Use, GAO-02-941 (Washington,
D.C.: Sept. 24, 2002), 3.
[32] Past GAO work provided information on the Florida Hurricane
Catastrophe Fund, California Earthquake Authority, and the Texas
Windstorm Insurance Association. See GAO-02-941 and GAO, Catastrophe
Insurance Risks: Status of Efforts to Securitize Natural Catastrophe
and Terrorism Risk, GAO-03-1033 (Washington, D.C.: Sept. 24, 2003).
[33] To maintain comparability with other data, GAO did not adjust
these data for changes in agricultural prices.
[34] There are three main catastrophe modeling firms: AIR Worldwide,
Risk Management Solutions, and EQECAT. Although many of the insurers we
interviewed use models from these firms, two of the eleven insurers
have developed their own catastrophe models.
[35] American Academy of Actuaries, Catastrophe Exposures and Insurance
Industry Catastrophe Management Practices (Washington, D.C.: American
Academy of Actuaries, June 10, 2001), Hyperlink,
http://www.actuary.org/pdf/casualty/catastrophe_061001.pdf (downloaded
Jan. 3, 2007), 10-12.
[36] Guy Carpenter, The World Catastrophe Reinsurance Market: Steep
Peaks Overshadow Plateaus (New York, N.Y.: Guy Carpenter, September
2006), Hyperlink,
http://www.guycarp.com/portal/extranet/insights/reports.html?vid=30
(downloaded Jan. 3, 2007).
[37] Insurance Information Institute, Catastrophes: Insurance Issues
(New York, N.Y.: Insurance Information Institute, November 2006),
Hyperlink, http://www.iii.org/media/hottopics/insurance/xxx/
(downloaded Jan. 3, 2007).
[38] Allianz Group and World Wildlife Fund, Climate Change and
Insurance: An Agenda for Action in the United States (New York, N.Y.:
Allianz Group and World Wildlife Fund, October 2006), Hyperlink,
http://www.allianz.com/en/allianz_group/sustainability/insight/studies_a
nd_reports/page1.html?hits=reports (downloaded Jan. 4, 2007).
[39] The report notes that these decisions were due, in part, to state
restrictions on rate increases that are designed to maintain insurance
prices that are affordable, but may not accurately reflect the true
potential for loss faced by the insured.
[40] Thirty-one states have FAIR plans, and six southern states have
state-sponsored wind insurance plans that pool resources from insurers
to cover the cost of coverage for their participants.
[41] Evan Mills and Eugene Lecomte, From Risk to Opportunity: How
Insurers Can Proactively and Profitably Manage Climate Change (Boston,
MA: Ceres, August 2006), Hyperlink,
http://www.ceres.org/pub/docs/Ceres_Insurance_Climate_%20Report_082206.p
df (downloaded Jan. 3, 2007), 34.
[42] Note that the federal government covers most, but not all,
payments in the event of loss under the FCIC--insurance providers also
share in the risk, as described in detail in appendix III.
[43] The Congress increased the NFIP's borrowing authority from $1.5
billion to approximately $20.8 billion in the wake of unprecedented
losses associated with the 2005 hurricane season.
[44] A detailed description of each program's risk management practices
can be found in appendixes II and III for the NFIP and FCIC,
respectively.
[45] Loss ratio, an indicator used to evaluate program performance, is
calculated by dividing claims paid by total premiums collected. A loss
ratio greater than 1.00 indicates that the program paid more in claims
than was collected in premiums
[46] The FCIC's average loss ratio from 1995 through 2005 was 0.91.
From 1981 through 1994, it was 1.47. See appendix III for more
information on the FCIC's performance.
[47] GAO, 21st Century Challenges: Reexamining the Base of the Federal
Government, GAO-05-325SP (Washington, D.C.: February 2005), 77.
[48] Pub. L. No. 90-448, 82 Stat. 573.
[49] Senate Committee on Banking and Currency, Insurance and Other
Programs for Financial Assistance to Flood Victims, 89th Cong., 2d
Sess., 1966, Committee Print.
[50] Pub. L. No. 93-234, 87 Stat. 975 (1973).
[51] Pub. L. No. 103-325, 108 Stat. 2255 (1994).
[52] The Bunning-Bereuter-Blumenauer Flood Insurance Reform Act of
2004, Pub. L. No. 108-264, 118 Stat. 712.
[53] In March 2003, FEMA and its approximately 2,500 staff became part
of the Department of Homeland Security (DHS). Most of FEMA--including
its Mitigation Division, which is responsible for administering the
NFIP--is now part of the department's Emergency Preparedness and
Response Directorate. However, FEMA retained its name and individual
identity within the department. Under a reorganization plan proposed by
the current Secretary of DHS, the Emergency Preparedness and Response
Directorate would be abolished, and FEMA would report directly to the
Undersecretary and Secretary of DHS.
[54] 42 U.S.C. § 4001(a)(4).
[55] Senate Committee on Banking and Currency, Insurance and Other
Programs for Financial Assistance to Flood Victims.
[56] GAO, Flood Insurance: Information on the Financial Condition of
the National Flood Insurance Program, GAO-01-992T (Washington, D.C.:
July 19, 2001).
[57] See 42 U.S.C. § 4016.
[58] Federal Crop Insurance Act, tit. V, 52 Stat. 72 (1938) (codified
as amended at 7 U.S.C. §§ 1501-1524).
[59] Pub. L. No. 96-365, 94 Stat. 1312 (1980).
[60] Pub. L. No. 103-354, 108 Stat. 3178 (1994).
[61] Pub. L. No. 104-127, 110 Stat. 888 (1996).
[62] Pub. L. No. 106-224, 114 Stat. 358 (2000).
[63] For example, if the claims paid in 1 year totaled $7.36 and the
insurance in force was $100, the loss-cost ratio is 7.36 percent. The
percentage represents the rate that would need to be charged per $100
of insurance coverage if total premiums are to equal the total claim
payments for that year. In this example, the 7.36 percent indicates
that a rate of $7.36 was required per $100 of insurance coverage sold.
[64] The excess of losses above the capped amount is pooled at the
state level and reallocated to the counties. According to FCIC, this
procedure is intended to reduce the variation of rates from one year to
the next.
[65] The surcharge is established by pooling the amount of insurance in
force and the claim payments for capped years with the highest loss-
cost ratios in each county that were not factored into the county
unloaded rates at the state level. These data are used to calculate a
statewide surcharge for catastrophic coverage (pooled claims payments
divided by pooled insurance in force). If the pooled losses at the
state level exceed five points, the excess is returned to the counties
and included in the county unloaded rate.
[66] Prevented planting factor adds a provision for losses due to crops
that were never planted because of external factors not directly
related to yield loss.
[67] The Federal Crop Insurance Reform Act of 1994 mandated
participation in the program to receive other commodity support
payments, although this requirement was rescinded in 1996.
GAO's Mission:
The Government Accountability Office, the audit, evaluation and
investigative arm of Congress, exists to support Congress in meeting
its constitutional responsibilities and to help improve the performance
and accountability of the federal government for the American people.
GAO examines the use of public funds; evaluates federal programs and
policies; and provides analyses, recommendations, and other assistance
to help Congress make informed oversight, policy, and funding
decisions. GAO's commitment to good government is reflected in its core
values of accountability, integrity, and reliability.
Obtaining Copies of GAO Reports and Testimony:
The fastest and easiest way to obtain copies of GAO documents at no
cost is through GAO's Web site (www.gao.gov). Each weekday, GAO posts
newly released reports, testimony, and correspondence on its Web site.
To have GAO e-mail you a list of newly posted products every afternoon,
go to www.gao.gov and select "Subscribe to Updates."
Order by Mail or Phone:
The first copy of each printed report is free. Additional copies are $2
each. A check or money order should be made out to the Superintendent
of Documents. GAO also accepts VISA and Mastercard. Orders for 100 or
more copies mailed to a single address are discounted 25 percent.
Orders should be sent to:
U.S. Government Accountability Office 441 G Street NW, Room LM
Washington, D.C. 20548:
To order by Phone: Voice: (202) 512-6000 TDD: (202) 512-2537 Fax: (202)
512-6061:
To Report Fraud, Waste, and Abuse in Federal Programs:
Contact:
Web site: www.gao.gov/fraudnet/fraudnet.htm E-mail: fraudnet@gao.gov
Automated answering system: (800) 424-5454 or (202) 512-7470:
Congressional Relations:
Gloria Jarmon, Managing Director, JarmonG@gao.gov (202) 512-4400 U.S.
Government Accountability Office, 441 G Street NW, Room 7125
Washington, D.C. 20548:
Public Affairs:
Paul Anderson, Managing Director, AndersonP1@gao.gov (202) 512-4800
U.S. Government Accountability Office, 441 G Street NW, Room 7149
Washington, D.C. 20548:
