[Economic Report of the President (2007)]
[Administration of George W. Bush]
[Online through the Government Printing Office, www.gpo.gov]

 
CHAPTER 5

Catastrophe Risk Insurance



Insurance plays a vital role in America's economy by helping
households and businesses manage risks. Individuals purchase
insurance so they can sleep well at night; they gain comfort from
the knowledge that they and their families are protected from
some of the adverse effects of future events beyond their control.
Businesses purchase insurance for much the same reason. It allows
them to reduce the uncertainty associated with future costs and
revenues, which enables them to plan for the future more
effectively. Today, one can purchase insurance protection against a
myriad of economic hazards, from poor health to motor vehicle
accidents to legal liability to lightning strikes.
Insuring economic losses arising from large-scale natural and
manmade catastrophes such as earthquakes, hurricanes, and terrorist
attacks poses special challenges for the insurance industry and for
Federal and State governments. This chapter examines the economics
of catastrophe risk insurance. It draws the following main
conclusions.

 In insurance markets, as in other markets, prices
affect the way people weigh costs and benefits. Insurance prices
that are artificially low can discourage people from adequately
protecting against future losses. For example, subsidized property
insurance prices may stimulate excessive building in high-risk
areas, potentially driving up future government disaster relief
spending.
 Government intervention in insurance markets can have
unintended consequences such as limiting the availability of
insurance offered by private firms.
 Private insurers manage catastrophe losses by being
selective about which risks to insure, by designing insurance
contracts to provide incentives for risk-reducing behavior, and by
charging prices that are high enough to enable them to diversify
risk over time or transfer risk to third parties. By adopting
private sector risk management and pricing practices, government
insurance programs could reduce the burden they impose on taxpayers
and minimize negative effects on private insurance markets.

The Economics of Catastrophe Risk Insurance

In the United States, insurance is provided through a variety of
private and public entities. Insurance companies owned by investors
or policyholders sell insurance in the private sector.
State-sponsored insurance pools have characteristics of both private
and public entities. They are typically owned by a group of private
insurers, but they are governed under charters that grant them
special rights and impose responsibilities not required of private
insurers. Finally, the Federal Government operates at least 135
different programs that provide insurance-like benefits to
individuals and businesses.
To understand how insurance works, imagine a large group of
homeowners scattered throughout the country, each of whom faces a
risk of property damage from a variety of identified hazards such
as fire or severe weather. The likelihood that any particular member
of the group will experience a loss is low, but the economic costs
to that individual, should a loss occur, are significant. Each
member of the group can reduce uncertainty about future economic
losses by agreeing to pool risk with other members. One way of
accomplishing this is through a mutual insurance agreement. At the
beginning of the year, each member agrees to make a payment,
called an insurance premium, into the pool. In exchange for their
premiums, members are allowed to file claims with the pool should
their houses incur damage from a covered hazard. Even if the
insurance pool has no other resources, as long as the total value
of premiums paid into the pool is at least as large as the value
of insured losses over the year, all property damage will be fully
covered. In this way, members of the pool gain security through
diversification. Because any member's losses are paid for with
premiums collected by all members, no member faces uncertainty about
how much he will have to pay to cover property damage in the coming
year.
The process of evaluating a risk exposure, determining whether or
not to insure it, and setting terms and conditions for any insurance
provided is called underwriting. Through underwriting, insurance
providers seek to tie the premiums charged for insurance policies to
the risks those policies cover. Effective underwriting serves an
important social function, because when insurance prices accurately
reflect underlying economic costs they can encourage a more
efficient  allocation of scarce resources. For example, suppose a
member of a coastal community must decide where to build a new home.
She may prefer to live as close to the ocean as possible, but a
home located nearer the ocean may be exposed to a higher risk of
damage from windstorms and flooding. If homeowners' insurance
premiums are appropriately risk sensitive, then she will need to
determine whether the benefits of living closer to the ocean are
worth the cost of higher insurance premiums.
Underwriting is critical to the efficient functioning of insurance
markets. In general, insurance markets function best under the
following conditions:

1.  Either all members of a pool face similar risks, or differences
in risks can be observed and incorporated in insurance premiums.
2.  Insurance does not dissuade those who are insured from avoiding
risks.
3.  The total value of insured losses for a pool can be forecast
with precision.



In many insurance markets, one or both of the first two conditions
may not hold. Violations of the third condition are a particular
feature of catastrophe-risk insurance markets. Through effective
underwriting, insurers can reduce, though perhaps not eliminate,
problems that arise when these conditions fail to hold.


Effective Underwriting Reduces Information Problems

Insurance markets may fail to work effectively when differences in
the risks faced by policyholders cannot be incorporated in insurance
premiums. To see why, consider again the example of homeowners
pooling risk. Suppose now that there are two types of homeowners:
those who live in coastal areas that are at relatively high risk
for windstorms and floods, and those who live in inland areas at
lower risk for these hazards. If all homeowners were charged the
same insurance premium, and if premiums were set equal to the
average loss rate for all homes, then homeowners in inland regions
would rightly feel that they were being overcharged. They face
less risk from windstorms and floods than owners in coastal regions,
yet they are asked to pay a premium equal to average losses for a
pool that includes houses in both regions. Owners living in
coastal areas would be attracted to the pool because it offers
insurance at a premium that does not reflect their homes' higher
risk. If the insurance policy were offered to all homeowners, a disproportionate share of those in coastal regions would accept
the policy, while a disproportionate share of those living inland
would seek insurance elsewhere or would choose to go without
insurance. As a result, the average loss for those who chose to
participate in the pool would be higher than the premium charged.
This example illustrates a general property of insurance contracts
which economists call adverse selection. When premiums do not
reflect differences in risk that are known to potential
policyholders, insurance pools tend to attract members who are at
greatest risk for the hazards covered. The solution to this problem
is to charge policyholders with different risk exposures different
premiums. In the example above, adverse selection could be avoided
if home-owners in inland areas were charged lower premiums than
those in coastal regions. Insurance providers generally try to set
premiums commensurate with risk, but this is not always possible.
In some cases it may simply be too costly for an insurance provider
to identify differences in risk, but, as discussed later in this
chapter, efforts by policymakers and insurance regulators to keep
premiums for some high-risk policyholders low can also play a role.
Inefficiencies can also arise when insurance discourages those who
are insured from taking actions to reduce potential losses. Consider
the incentives faced by a homeowner thinking about how best to
prepare for future windstorms. Many homeowners can reduce the
damage caused by windstorms by installing storm shutters, but storm
shutters are costly. If a homeowner is fully insured against the
economic losses arising from future windstorms, she may


be less likely to purchase shutters. The tendency of those who are
insured to work less hard to avoid losses is called moral hazard.
Insurance providers are well aware of the potential for moral
hazard, and they attempt to address it through effective
underwriting. Many insurance policies only cover losses in excess
of a specified amount called a deductible, or they require that
policyholders pay a fixed share of any losses incurred. By insuring
some, but not all, economic losses, these types of policies
strengthen policy-holders' incentives to work to reduce the risks
they face. Insurers may also require that specific action be taken
as a precondition for receiving coverage, or they might provide
pricing incentives for risk-reducing investments. For example, an
insurer might refuse to cover windstorm risks for homes without
storm shutters, or it might charge those homeowners a higher
premium.

Catastrophe Losses Are Difficult to Forecast

Adverse selection and moral hazard problems are common in many
insurance markets. Catastrophe risk insurers face an additional
challenge, which arises from the fact that the total value of
losses for a pool of insured properties or individuals is often
exceptionally difficult to predict.
Forecasting annual losses from hazards like automobile accidents
that only affect one or two members of a pool at a time is much
easier than forecasting losses from large-scale catastrophes such
as floods, hurricanes, or terrorist attacks. When the losses
incurred by individual members of an insurance pool are more or
less independent of one another, the average loss rate per policy
is likely to be stable over time. Chart 5-1 illustrates this point
by showing the annual nationwide accident rate per 100,000
registered passenger cars. While the accident rate has gradually
declined over the past 15 years, it changes relatively little
from year to year. It is difficult to predict whether any particular
vehicle will be involved in an accident, but based on the data
presented we can forecast with high confidence that about 4.5
percent of all passenger cars will be involved in some kind of
accident over the next year. Because large-scale catastrophes have
the potential to affect many members of an insurance pool
simultaneously, spreading risk across a large number of members
may not be sufficient to ensure that average losses per policy are
stable over time. Compare Chart 5-1 with Chart 5-2. Chart 5-2
reports the number of loss claims filed per 100,000 homes and
businesses insured for flood losses under the Federal Emergency
Management Agency's National Flood Insurance Program (NFIP). Flood
losses are not independent of one another; a single flood event
can damage hundreds or even thousands of properties. Even though
the NFIP insures a pool of millions of properties, the average
loss rate per policy varies considerably from year to year.





In some catastrophe-risk insurance markets, forecast accuracy also
suffers from a lack of relevant historical data and experience.
This is a particular problem when catastrophes are rare, and when
the character of those events is likely to change over time. For
example, U.S. commercial property and casualty insurers had almost
no experience forecasting losses from large-scale terrorist attacks
prior to September 11, 2001. A recent report by the President's
Working Group on Financial Markets on the availability and
affordability of insurance for terrorism risk found that while
modeling of terrorism risk has improved since 2001, insurers
continue to have limited confidence in the models they use for
evaluating this risk exposure.
When annual losses for a pool can be forecast with reasonably high
precision, it is relatively easy for an insurance provider to manage
risk. As long as its underwriting procedures ensure that the average
premium paid by members of the pool is at least as large as the
average loss rate per member, it is likely that in any given year
total premium revenues for the pool will be sufficient to pay all
claims. If, as in our automobile accident example, losses are
independent across members of a pool, increasing the size of the
pool actually makes it easier for an insurer to manage risk,
because the more members that are included in the pool, the more
stable will be the average loss rate per member.
Losses from catastrophes are not independent across exposures, and
therefore they are much more difficult to manage. A severe
hurricane, for example, can cause damage over tens of thousands
of square miles, so even if an insurer provides windstorm coverage
for properties scattered throughout a state, average losses per
property are likely to be exceptionally high in hurricane years.
Since catastrophes are infrequent but costly, annual premium
revenues for a pool of exposures that exceed the value of claims
in most years may not be sufficient to pay all claims in those
rare years when a severe event occurs. Insurance providers work to
address this problem by pooling risk across time or by diversifying
the risk exposure more broadly by sharing it with other insurers.


Managing Catastrophe Losses

One way to manage the financial risk of insuring catastrophe hazards
is to retain a portion of excess premium revenues collected in years
when losses are low to pay claims in years when catastrophes
generate large losses. Equity capital set aside to pay potential
claims is called surplus. In practice, building surplus large enough
to pay catastrophe losses can be difficult for private insurance
companies. Owners of insurance companies expect to earn a market
rate of return on their equity investments, including equity held
as surplus to cover future claims. Moreover, income flowing from
insurance company assets is subject to corporate income tax that
effectively adds to the cost of accumulating and holding surplus.


An alternative to using surplus to cover catastrophe losses is to
transfer risk to third parties. Some insurers transfer risk directly
to capital market partici-pants such as hedge funds and
institutional investors (Box 5-1). More commonly, insurers
negotiate risk-sharing agreements with specialized insurance
companies called reinsurers. Reinsurers are internationally
diversified companies that make a business of selling insurance to
primary insurers. In a typical reinsurance arrangement, a primary
insurer pays a fee to a reinsurance company that agrees to cover
some of the insurer's costs in the event that claims exceed a
prespecified threshold. In essence, reinsurance arrangements work
much like other types of insurance. Through reinsurance a primary
insurer subject to the risk of high claims caused by a catastrophe
can pool its risk with other primary insurers that are exposed to
different hazards. As with other types of insurance, problems of
adverse selection and moral hazard can impede the efficient
functioning of reinsurance markets.

-------------------------------------------------------------------
Box 5-1: Catastrophe Bonds and Sidecars-Accessing Financial Markets
to Better Manage Catastrophe Risks

Though reinsurance agreements between primary insurers and
specialized reinsurance companies remain the most popular method
for transferring and pooling risks posed by large-scale
catastrophes, the capital available to reinsurers is only a tiny
fraction of the total capital invested
in financial markets. By one estimate, reinsurance companies
worldwide had accumulated about $400 billion in shareholder funds
by year-end 2005, which is only about 1 percent of the market
capitalization of the world's public equity markets. To spread
catastrophe risks more broadly, financial markets have developed
mechanisms to allow investors who do not directly hold shares in
insurance companies to assume some of the catastrophe risk
exposure of primary insurers or reinsurers in exchange for an
appropriate investment return. Two notable examples are catastrophe
bonds and ``sidecars.''
Catastrophe bonds (CAT bonds), also called ``acts of God'' bonds, are
risk-linked securities that offer a return to investors similar to
that on high-yield corporate junk bonds. In a typical CAT bond
transaction, a firm that wants to transfer some risk to outside
investors issues a bond and invests the proceeds in safe securities.
If a specified catastrophe event occurs, the proceeds from the bond
issue are released to the issuer. If no event occurs during the term
of the bond, the principal is returned to investors. Payouts from
CAT bonds are often tied to industry-wide loss estimates or defined
catastrophe events such as whether or not a hurricane makes landfall
on a particular stretch of coastline. Because these types of events
are presumably beyond the continued on the next page
control of the bond issuer, investors are protected from moral
hazard. A drawback of these types of CAT bonds, however, is that
they do not protect the issuer against all possible catastrophe
losses. For example, an insurer that issues a bond with a payout
tied to a hurricane event could be exposed to large losses from a
tropical storm that does not meet the definition of a hurricane.
The market for CAT bonds has grown rapidly over the past decade,
though the value of bonds outstanding remains small relative to
the value of insured losses in recent catastrophe events. About
$4.9 billion in CAT bond capital was outstanding as of year-end
2005, a 21 percent increase over the 2004 level.
Sidecars provide an increasingly popular alternative to CAT bonds.
A sidecar is a special-purpose financial entity, usually designed
to last 2 to 3 years. Under a sidecar arrangement, a group of
investors partners with an existing reinsurance company: the
investors provide the necessary funds for deployment and the
reinsurance company contributes its infrastructure, business
relationships, and the skills of its staff. Sidecar investors
receive a portion of the reinsurance company's premium revenue
from a particular reinsurance contract or line of busi-ness, and
the reinsurer gains access to the investors' capital to cover
potential catastrophe losses. Through sidecars, investors can
decide to assume particular catastrophe risks without being exposed
to all of the risks covered by a given reinsurance company.
Sidecars have helped Bermuda-based reinsurance companies to expand
their capacity to cover catastrophe risk exposures in the United
States despite incurring significant losses in 2005. About $2.5
billion in capital was reportedly raised through sidecars
organized with Bermuda reinsurers from December 2005 to June
2006.
Through CAT bonds, sidecars, and other innovative financing
mechanisms, insurers and private investors are finding new ways to
spread the risks posed by large-scale catastrophes. These financing
mechanisms currently contribute only a relatively small share of
the total capital available to cover catastrophe losses, but the
volume of capital they have raised has grown rapidly in recent
years. It is likely that as these markets mature, the base of
investors willing to bear some catastrophe risk will continue to
expand, ultimately lowering the cost of insuring catastrophe risks.
--------------------------------------------------------------------

What happens if an insurance provider lacks the resources to pay
claims following a catastrophe? Private-sector insurance companies
that cannot afford to pay claims are usually forced into
receivership. In contrast, many government-sponsored insurers can
raise additional funds to pay claims after an event has occurred.
Government-sponsored insurance programs often do not face the same
financial constraints as private insurers because they have special
rights to compel third parties such as taxpayers or private insurers
to bear a portion of their financial risk. The NFIP, for example, is
authorized by Congress to borrow from the U.S. Treasury, which
increases taxpayer liabilities, and the Federal Government's
terrorism-risk insurance program and several State-sponsored
catastrophe insurance providers are empowered to levy surcharges on
policies sold by private insurers.


Federal Catastrophe Insurance Programs

In 1803, Congress passed a law granting the victims of a fire in
Portsmouth, New Hampshire, extra time to repay certain debts owed
to the Federal Government. Though the Federal Government has
assisted Americans harmed by disasters throughout the Nation's
history, prior to the mid-twentieth century aid was generally
provided on an ad hoc basis; a disaster would strike and Congress
would then determine whether and to what extent Federal aid would
be provided. Acts of Congress passed in 1947 and 1950 regularized
the process by which the Federal Government extends assistance to
disaster-affected communities and additional legislation enacted
since then has clarified and expanded the Government's role in
disaster relief.
One problem with a variety of government relief efforts is that they
can make it more difficult for private insurers to sell policies
for some catastrophe hazards at prices commensurate with underlying
risks. People have less incentive to pay sometimes high insurance
premiums if they expect to receive aid from the government when a
catastrophe strikes. Policymakers have sought to address this moral
hazard problem in several different ways. The Federal Government
provides insurance coverage for certain catastrophe hazards, often
at prices lower than those that would be charged by private
insurers. In addition, in some cases the Government requires that
individuals purchase insurance policies or mandates that private
insurers offer policies for sale.


The National Flood Insurance Program

The National Flood Insurance Program (NFIP) was established in
1968 to make flood insurance more widely available to homeowners
and businesses, to encourage local communities to prepare better
for flood hazards, and to reduce reliance on direct Federal
disaster relief following floods. The NFIP
currently provides flood insurance for 5.3 million policyholders
nationwide, many of whom might not be able to obtain coverage
without the program. Residential and commercial property owners in
some 20,000 participating communities are eligible to purchase
flood insurance policies under the program. Homeowners with
mortgages issued by federally regulated lenders on property in
communities identified to be in flood hazard areas are required
to purchase flood insurance on their dwellings. Property owners
can purchase policies either directly from the Federal Government
or, more commonly, through local insurance companies who sell
NFIP policies under their own name but pass their risk on to the
Government. Whether policies are sold directly by the Federal
Government or by insurance companies, the NFIP receives premium
payments for the policies and bears all financial risks associated
with the insurance they provide. The program is administered by
the Federal Emergency Management Agency (FEMA).
FEMA relies on Flood Insurance Rate Maps (FIRMs) when underwriting
flood insurance. These maps identify areas within a community that
have at least a 1-percent chance per year of being inundated by
high water. These areas are called 100-year floodplains. Federal
flood insurance is only made available in local communities that
agree to adopt zoning ordinances, building codes, and other planning
measures designed to reduce future damage caused by floods. For
example, communities must require that new buildings be elevated
above the level that flood waters are expected to reach on average
once per 100 years. According to FEMA, buildings that meet its
floodplain management standards suffer 80 percent less damage from
floods each year than those that do not. Not all structures insured
under the NFIP meet these standards, however; structures completed
prior to a community's decision to participate in the program or
prior to the publication of a community's FIRM are eligible for
insurance under the program even if they do not meet FEMA standards.
The NFIP charges different premiums for different properties.
A structure built or substantially renovated after 1974 or after a
community's FIRM was completed (whichever is later) is charged an
actuarially fair annual premium equal to an estimate of expected
annual claims under the property's flood insurance policy.
Policyholders who pay actuarially fair premiums year after year
should, in the long run, end up paying premiums that are just
sufficient to cover their claims on average. About one-quarter of
NFIP policies cover properties built prior to 1974 or prior to the
publication of a community's FIRM. By law, these ``pre-FIRM''
properties are charged subsidized premiums. Pre-FIRM properties are
much less likely to comply with modern flood risk mitigation
standards since most were built before such standards were widely
applied. Because of their higher risk, pre-FIRM properties are
assessed higher premiums on average than newer properties, but
even these higher premiums are not adequate to cover expected
losses. On average, premiums for pre-FIRM properties represent
only about 40 percent of those properties' actuarially fair rates.
Not surprisingly, the NFIP pricing scheme has led to serious
adverse selection and moral hazard problems. On the one hand, FEMA
estimates that one-half to two-thirds of structures in floodplains
do not carry flood insurance. On the other hand, some exceptionally
high-risk properties continue to receive NFIP coverage at subsidized
rates even though they have been damaged by floods multiple times
since entering the program. Some 50,644 properties insured by the
NFIP as of September 30, 2004 had incurred flood damage resulting
in claims of at least $1,000 more than once during a 10-year
period. While these properties only represented about 1 percent of
all structures then insured under the program, repetitive-loss
properties have historically accounted for 38 percent of all
program claims payments. Amendments to the Flood Insurance Act
passed in 2004 authorized a pilot program to remove some of the
most severe repetitive-loss properties from the NFIP insurance
roll by allowing FEMA to fund work to elevate or relocate some
of them or, in extreme cases, to purchase and demolish them.
The NFIP illustrates how underwriting standards can either enhance
or impede loss mitigation. By providing coverage only in
communities that agree to adopt flood-risk mitigation measures,
the NFIP may have induced some communities to take steps that FEMA
credits with reducing flood damage by an average $1.2 billion
annually. At the same time, by providing insurance to pre-FIRM
properties at less than actuarially fair rates, the program may
have discouraged some policyholders from relocating or renovating
structures at high risk for flood damage. The availability of
flood insurance has lowered the risk to banks of financing
real-estate investment in locations vulnerable to flood losses. As
a result, it is not clear whether the NFIP has reduced the size of
Federal appropriations for flood disaster relief as intended.
Demand for Federal disaster aid may arguably be higher than it
would have been had the NFIP not facilitated development in
high-risk areas.
Chart 5-3 shows that since 1986 NFIP premiums exceeded annual losses
in most years, but were woefully inadequate to cover losses from
Hurricanes Katrina, Rita, and Wilma in 2005. The 2005 hurricanes
resulted in about $16.3 billion in NFIP program claims, some of which
were not paid until 2006. Even so, claims paid in 2005 exceeded
premiums collected in that year by a factor of nearly six to one.
Unlike private sector insurers, who would need to accumulate
surplus or purchase reinsurance to pay claims in excess of premiums,
the NFIP is permitted to borrow from the Federal Government. As of
August 2005, just before Hurricane Katrina struck, the NFIP had
accumulated a relatively modest $300 million in debt owed to the
U.S. Treasury, but the program will need to borrow an additional
$21.2 billion to pay claims  filed in 2005. Though the NFIP is
supposed to repay this debt using
future premium revenue, it is unlikely that this will be possible.
The Congressional Budget Office estimates that by 2007 the interest
on NFIP debt will grow to about $1 billion annually, which is about
40 percent of the projected annual premium revenue. Even if future
hurricane seasons are milder than those experienced in recent years,
projected premiums are not expected to be large enough to cover both
the interest on the outstanding debt and the projected future
claims. The NFIP's current dire financial situation amply
demonstrates that in insurance, as elsewhere, there is no free
lunch. Annual premium revenue from the NFIP was able to cover losses
in most of the program's recent history, but the subsidized insurance
program exposed the American taxpayers to a huge potential financial
liability which became an actual liability in 2005.


Terrorism and War-Risk Insurance Programs

The Federal Government provided billions of dollars in disaster
assistance following the September 11, 2001 terrorist attacks on New
York and Washington, DC, including about $4 billion in aid to the
airline industry and about $20 billion in aid to the New York City
area. To date, about $36 billion





in loss claims have been paid by private insurers. Though insured
losses represented only a fraction of the total economic costs of
the September 11 attacks, they were far greater than those arising
from any prior terrorist event.
Following September 11, commercial property and casualty insurers
reevaluated their policyholders' exposure to risk from possible
future attacks. Many insurers canceled policies, began explicitly
excluding coverage for terrorist attacks from new policies, or
increased premiums charged to policyholders. In response to what was
believed to be a temporary contraction in the supply of insurance
available for terrorism risk, the Administration and Congress
undertook measures to ensure that the airline and commercial real
estate sectors would not be adversely affected.
Less than two weeks after the September 11 attacks, the Federal
Aviation Administration (FAA) began selling insurance policies
directly to U.S. airlines to cover third-party liability (e.g.,
harm to individuals or property on the ground) arising from acts
of war or terrorism, and in November of 2002 the Homeland Security
Act expanded this program to provide insurance coverage for loss
of aircraft and airline passenger liability as well. The program
has been reauthorized several times since its inception and it
remains in effect today. As of October 1, 2006, policies under this
program provided 75 airlines with insurance coverage for potential
losses ranging from $100 million to $4 billion each.
The Terrorism Risk Insurance Act (TRIA) passed in November of 2002
established a second, much broader, Federal program to encourage
private-sector commercial property and casualty insurers to provide
terrorism risk coverage. The program was originally designed to
expire after three years, but in 2005 Congress elected to extend
the program with some modifications through 2007.
TRIA has two main components. First, it mandates that insurance
companies that sell commercial property and casualty insurance make
available to customers policies that do not explicitly exclude
coverage for losses caused by acts of terrorism. Insurers may
exclude losses on other grounds, however, so not all losses arising
from terrorist attacks must be covered. According to the President's
Working Group on Financial Markets, commercial insurance policies
generally do not cover losses arising from chemical, nuclear,
biological, and radiological events, whether or not these events
are caused by acts of terrorism. Second, TRIA authorizes the Treasury
Department to provide reinsurance to cover a portion of insurance
loss claims arising from certified acts of international terrorism
against U.S. targets. Under the reinsurance program, a primary
insurer must cover 100 percent of its loss claims up to a specified
deductible. The Federal Government then pays a fixed share of losses
in excess of the deductible. For 2007 an insurance company is
required to cover all losses up to 20 percent of its prior year's
premiums on qualifying


lines of business and 15 percent of losses above this deductible.
TRIA imposes a cap of $100 billion on total insurer losses from
terrorist attacks. Under the statute, Congress would determine the
procedures to govern any payments for losses beyond $100 billion
in separate legislation.
Since 2001, no claims have been filed under either the FAA's aviation
war-risk insurance program or the Treasury Department's
terrorism-risk reinsurance program, but, like the NFIP, both of
these programs expose U.S. taxpayers to large potential losses.
Because they were intended to be temporary, neither program is
designed to ensure that premiums will be sufficient to pay future
claims. Premium revenue collected under the aviation war-risk
program is subject to a cap mandated by Congress. As a result,
premiums charged by the FAA are significantly lower than those
that would be charged for comparable policies sold by private-sector
aviation insurers. Airlines pay a total of about $160 million in
premiums to the FAA each year; by one estimate, without the program
these airlines would need to pay $500 million annually in premiums
to private insurers. TRIA does not require property and casualty
insurers to pay any premiums for the reinsurance protection they
receive. Instead, claims under the program are expected to be
paid with Federal outlays and then recouped, after the fact,
through surcharges levied on future premiums for property and
casualty insurance policies. Given that the program was established
in part to address problems arising from high insurance premiums
following the September 11, 2001 attacks, there are real questions
as to whether surcharges would be set high enough to recoup
expenditures following a future terrorist attack. Any surcharges
would likely be spread over several years to reduce the impact on
premiums, and since the Treasury Department is only required by
law to recoup up to $27.5 billion, there is no guarantee that the
full costs of the program would ultimately be recovered.


State Property Insurance Markets

Although the Federal Government is actively involved in insuring
risks from floods and terrorist attacks, most homeowners and
businesses look first to their local property insurers to obtain
financial protection against a variety of hazards including
potential catastrophes. State governments are responsible for
regulating insurance markets. Though laws differ from state to
state, all states' insurance regulators exercise some control over
who is permitted to sell insurance, what terms and conditions can
be attached to insurance policies, and how much insurers can
charge. Insurance regulations are intended to protect consumers
who may have difficulty evaluating complex insurance contracts and
to ensure that insurers maintain sufficient financial resources
to pay future claims. While regulation plays an important role in
protecting


consumers from fraud and poor risk management practices, poorly
conceived and executed regulation can create long-term problems
for the operation of state catastrophe-risk insurance markets.
Every state regulates property insurance premiums charged to
homeowners and small businesses. Many states require that premiums
be approved in advance by regulators. Others allow insurance
regulators to review existing price schedules and empower regulators
to force companies to reimburse policyholders when premiums are
found to be excessive. Rate regulations can make it difficult for
insurance companies to set premiums that accurately reflect
available information about risks, which can exacerbate moral
hazard and adverse selection problems. In some states the rate
review and approval process can take many months, so insurers
cannot rapidly adjust premiums when new information becomes
available. The rate review process may also discourage insurance
companies from proposing complex pricing plans which, though
difficult to explain and justify to state rate boards, more
accurately reflect detailed information about the risks associated
with individual insurance policies.
Efforts by regulators to keep property insurance prices
artificially low can make it difficult for individuals and
businesses to obtain insurance on private markets at any price.
To ensure that they will be able to pay claims after a catastrophe,
private insurers need to set premiums high enough to enable them
to build surplus or transfer risk to reinsurers. If regulators do
not allow insurers to charge rates sufficient to accomplish these
tasks, the insurers will be discouraged from taking on catastrophe
risks. They may choose to sell insurance only in areas at low
risk for catastrophe hazards, or they may seek to exclude coverage
for such hazards under the terms of the property insurance policies
they offer. Regulation can also deter insurers from competing for
customers, thereby reducing the range and quality of insurance
options available.
Many states that face risks from hurricanes or earthquakes have
established special entities to provide insurance to those who cannot
obtain coverage from private insurers. In 1996, California
established a quasi-public company, the California Earthquake
Authority, to sell earthquake insurance policies to California
residents, backed by funds contributed by a number of private
insurers operating in the state. Several states maintain residual
pools to cover windstorm risks. These pools operate like traditional
insurance companies, but they are required to sell policies to
property owners in high-risk coastal areas and they are empowered
to levy surcharges on primary insurers operating in a state.

Some state-sponsored insurance programs use complicated procedures
for setting premiums, and many claim to charge premiums that are
actuarially fair, but they all have one thing in common: they
provide insurance only to policyholders who either will not, or
cannot, obtain insurance from the

---------------------------------------------------------------------
Box 5-2: Gulf Coast Property Insurance Markets After Hurricanes
Katrina, Rita, and Wilma 2005 was a terrible year for communities
located along the U.S. Gulf Coast. Hurricane Katrina devastated a
land area the size of Great Britain and displaced more than 270,000
people. The total value of property damage and business interruption
caused by Hurricane Katrina has been estimated at $135 billion.
Hurricane Katrina was followed a few weeks later by Hurricane Rita,
which caused an estimated $15 billion in damage, and Hurricane Wilma,
which caused an estimated $20 billion in damage. The President and
Congress responded by appropriating about $110 billion for disaster
relief and recovery aid to affected communities. Property insurers
have also played an important role in recovery efforts by paying
billions of dollars of loss claims, but there are concerns that
rising insurance premiums for coastal properties may be a barrier
to redevelopment. The response of property insurance markets to
the unprecedented losses caused by the 2005 hurricane season
underscores the role of effective underwriting in managing
catastrophe risks.
Hurricanes Katrina, Rita, and Wilma resulted in an estimated $57
billion in insured property damages, not including claims filed
with the National Flood Insurance Program. Despite bearing enormous
losses, most private-sector primary insurers operating in the Gulf
Coast emerged from the 2005 hurricane season in reasonably sound
financial condition. At least four primary insurers failed as a
result of the 2005 storms, but the share of property and casualty
insurers listed as financially impaired by a major insurance
company rating agency actually dropped to a 25-year low while the
aggregate value of surplus available to insurers for paying future
claims increased. Primary insurers fared well as a group in part
because they had transferred a significant share of their
catastrophe risk exposure to reinsurers. According to one
industry association, reinsurance covered about 60 percent of 2005
insured hurricane losses.
Though the U.S. property and casualty insurance sector as a whole
remains healthy, property insurance markets in several coastal
states are under stress. Information collected during the 2004 and
2005 hurricane seasons revealed deficiencies in industry-standard
catastrophe risk models used in underwriting property insurance.
These models are now being adapted to reflect expectations of
more violent hurricane seasons, revised analysis of the costs of
repairing property damage following major catastrophes, new
findings about the effects of hurricane-generated storm surges,
and other factors. As a result, primary insurers and reinsurers
are increasing their estimates of probable losses on windstorm
policies in areas at risk for hurricanes. A leading
catastrophe-risk modeling firm reports that revised forecasts of the
severity of Atlantic hurricane seasons alone will increase estimates
of loss rates from future hurricanes in the Gulf Coast and
southeastern U.S. by 50 percent.

As assessments of the potential costs of future hurricanes have
increased, primary insurers and reinsurers have sought to limit
their exposure to windstorm hazards and increase the premiums
charged for insuring this hazard. Reinsurance companies, many of
whom lost capital in 2005 to hurricane-related claims, have
significantly increased premiums. Unlike reinsurance premiums,
premiums charged by primary insurers for homeowners' and commercial
property policies are regulated by state insurance commissions.
Primary insurers have petitioned state regulators to allow them
to raise premiums to cover rising reinsurance costs and to more
closely reflect new information on the risks posed by windstorms.
Where possible, some insurers have also attempted to reduce their
exposure to windstorm hazards by refusing to renew existing
policies in high-risk areas or by adding conditions to policies
that exclude coverage of windstorm damage. In several states,
government-sponsored insurance programs that are required to provide
windstorm coverage to property owners who are unable to obtain
insurance through the private sector have grown dramatically.
Recent developments in coastal property insurance markets have
the potential to discourage some investment in areas at high risk
for hurricanes, since property owners in these areas will likely
have to pay higher insurance premiums or bear greater risk than in
the past. For this reason, some have argued that Federal and State
governments should take action to ensure that insurance for
windstorm coverage in hurricane-prone regions is widely available
and that the premiums charged for this insurance are relatively low.
However, as discussed in the text, efforts to keep premiums for
windstorm insurance artificially low may discourage property owners
from taking action to lessen future windstorm losses while
potentially encouraging excessive development in high-risk areas.
private market. These programs tend to attract exactly those members
whose high risk makes them unattractive to private insurers.
For example, in some states, residual pools are the main providers
of windstorm insurance for homeowners in coastal areas exposed to
high risk from hurricanes.
In recent years a number of state-sponsored insurance programs have
had difficulty paying claims following major catastrophes. Different
states have dealt with this problem in different ways. A few states
have used government money to provide new funds for insolvent
programs, thereby passing the cost


of covering losses on to taxpayers. More commonly, states have levied
surcharges on premiums for policies sold by private insurers. This
approach effectively forces property owners in relatively low-risk
areas who can obtain insurance from private providers to pay higher
premiums to cover insured losses for property owners in higher risk
areas who obtain insurance through the residual pool. By effectively
raising the cost of insurance in the private market, these
surcharges may actually encourage more property owners to seek
insurance from the residual pool so that the pool is exposed to even
higher losses the next time a catastrophe strikes.
Since people consider the cost of property insurance when deciding
where to live and conduct business, the use of rate regulations or
state-sponsored insurance programs to keep property insurance prices
in high-risk areas artificially low can have significant negative
consequences. All else equal, commercial and residential development
will tend to be greater in those areas where insurance prices are
lower. As a result, artificially low premiums for catastrophe risk
insurance can lead to excessive development in catastrophe-prone
areas, putting lives and property in harm's way.


Conclusion

All insurance markets are susceptible to problems arising from
adverse selection and moral hazard, but insurers of catastrophe
risks must also deal with the fact that total insured losses are
difficult to predict and are potentially quite large. While it
may not be possible to eliminate these problems, their effects
can be moderated through prudent underwriting. Adverse selection
and moral hazard problems can be lessened by being selective about
which risks to insure, by setting premiums to match observable
differences in risk, and by requiring policyholders to bear a share
of the financial risk posed by the hazards they are insured against.
Insurance providers deal with uncertain losses by charging premiums
that are high enough to enable them to build surplus and/or transfer
excess risk to third parties such as reinsurers.
Regulations that constrain private insurers' underwriting
flexibility can undermine their ability to provide insurance
coverage for catastrophe risks. Government-sponsored insurance
programs that can borrow from the U.S. Treasury or levy surcharges
to pay claims after a catastrophe has occurred do not face the same
financial constraints as private insurers. Nonetheless, government
programs that do not apply prudent underwriting standards expose
taxpayers to large liabilities.
Effective insurance underwriting serves an important social
function by tying the premiums and terms of insurance policies to
the risks covered. When insurance prices reflect underlying economic
costs they can encourage a more efficient allocation of resources.
Efforts to keep premiums for insurance against catastrophe hazards
artificially low, whether through regulation or through subsidized
government programs, can encourage excessively risky behavior on the
part of those who might be affected by future catastrophes.