Flood Insurance: Financial Resources May Not Be Sufficient to Meet Future
Expected Losses (Letter Report, 03/21/94, GAO/RCED-94-80).

The National Flood Insurance Program, a key component of federal flood
disaster relief, was always intended to be subsidized.  Yet overall
premium income for the program is not enough to build reserves that can
cover anticipated flood losses.  Forty-one percent of the program's
policies are subsidized, and it is inevitable that claims losses and
program expenses will at some point exceed program funds.  In the event
of a catastrophe, not even the Federal Emergency Management Agency's
borrowing authority would be enough to cover potential claims.  Raising
premiums would improve the program's financial health but could lead to
canceled policies, creating a future burden on other flood relief
programs.  On the other hand, greater program participation by property
owners, although likely to cut the cost of other federal disaster
assistance programs, would place a greater financial burden on the
National Flood Insurance Program because of the need to cover additional
subsidized properties.

--------------------------- Indexing Terms -----------------------------

 REPORTNUM:  RCED-94-80
     TITLE:  Flood Insurance: Financial Resources May Not Be Sufficient 
             to Meet Future Expected Losses
      DATE:  03/21/94
   SUBJECT:  Program evaluation
             Flood insurance
             Property damages
             Revolving funds
             Emergency preparedness
             Disaster relief aid
             Property losses
             Insurance premiums
             Subsidies
             Financial management systems
IDENTIFIER:  FEMA National Flood Insurance Program
             
**************************************************************************
* This file contains an ASCII representation of the text of a GAO        *
* report.  Delineations within the text indicating chapter titles,       *
* headings, and bullets are preserved.  Major divisions and subdivisions *
* of the text, such as Chapters, Sections, and Appendixes, are           *
* identified by double and single lines.  The numbers on the right end   *
* of these lines indicate the position of each of the subsections in the *
* document outline.  These numbers do NOT correspond with the page       *
* numbers of the printed product.                                        *
*                                                                        *
* No attempt has been made to display graphic images, although figure    *
* captions are reproduced. Tables are included, but may not resemble     *
* those in the printed version.                                          *
*                                                                        *
* A printed copy of this report may be obtained from the GAO Document    *
* Distribution Facility by calling (202) 512-6000, by faxing your        *
* request to (301) 258-4066, or by writing to P.O. Box 6015,             *
* Gaithersburg, MD 20884-6015. We are unable to accept electronic orders *
* for printed documents at this time.                                    *
**************************************************************************


Cover
================================================================ COVER


Report to Congressional Requesters

March 1994

FLOOD INSURANCE - FINANCIAL
RESOURCES MAY NOT BE SUFFICIENT TO
MEET FUTURE EXPECTED LOSSES

GAO/RCED-94-80

National Flood Insurance


Abbreviations
=============================================================== ABBREV

  BFE - base flood elevation
  BFEWH - base flood elevation with wave height
  DED - deductible offset
  DELV - damage by elevation
  EXLOSS - expense item factor
  FEMA - Federal Emergency Management Agency
  FIA - Federal Insurance Administration
  FIRM - flood insurance rate map
  GAO - General Accounting Office
  LADJ - loss adjustment factor
  OIG - Office of Inspection General
  OMB - Office of Management and Budget
  PELV - probability of elevation
  SFHA - special flood hazard area
  UINS - underinsurance factor

Letter
=============================================================== LETTER


B-255740

March 21, 1994

The Honorable Donald W.  Riegle, Jr.
Chairman, Committee on Banking, Housing,
 and Urban Affairs
United States Senate

The Honorable Paul S.  Sarbanes
Chairman, Subcommittee on Housing
 and Urban Affairs
Committee on Banking, Housing,
 and Urban Affairs
United States Senate

The Honorable John F.  Kerry
United States Senate

This report responds to your request for information on the financial
condition of the National Flood Insurance Program administered by the
Federal Emergency Management Agency's (FEMA) Federal Insurance
Administration.  The program, along with low-interest loans provided
by the Small Business Administration and individual and family grants
provided by FEMA, is a major component of the federal government's
efforts to provide flood-related disaster assistance.  Floods during
fiscal year 1993, including last summer's flood in the Midwest, have
resulted in the payment of hundreds of millions of dollars in federal
flood insurance claims that have drained the cash reserves of the
program.  Members of Congress and the public have raised concerns
about whether the program has sufficient financial resources to meet
its current and potential future obligations. 

Prior to the inception of the flood insurance program in 1968, flood
insurance was generally not available from private insurance
companies.  The National Flood Insurance Act of 1968 (P.L.  90-448)
established the program to identify flood-prone areas, make flood
insurance available to property owners living in communities that
joined the program, encourage floodplain management efforts to
mitigate flood hazards, and reduce federal expenditures on disaster
assistance.  Flood insurance rate maps (FIRM) were prepared to
identify special flood hazard areas.  In order for a community to
join the program, any structures built within a special flood hazard
area after the FIRM was completed were required to be built to the
program's building standards that are aimed at minimizing flood
losses.  Owners of these post-FIRM structures pay actuarial rates for
national flood insurance.\1 By contrast, subsidized insurance rates
are available for owners of older, generally less flood-worthy
pre-FIRM structures. 

This report provides information on the (1) actuarial soundness of
the program, (2) potential financial impacts of increasing subsidized
flood insurance rates and enhancing program participation, and (3)
procedures used to set the program's insurance rates.  In addition,
as agreed with your offices, appendix I updates our review\2 of
FEMA's actions on its financial management problems addressed in
audits of the fund that were prepared by FEMA's Office of Inspector
General. 


--------------------
\1 An actuarial rate is risk-based because it considers the financial
risk to the insurer in issuing an insurance policy.  For the entire
program to be actuarially sound, the overall revenues from insurance
premiums would need to be sufficient to cover expected claims losses
and program expenses. 

\2 We discussed both the actuarial soundness and financial management
issues in our testimony before the Subcommittee on Housing and Urban
Affairs, Senate Committee on Banking, Housing, and Urban Affairs. 
Flood Insurance:  Information on Various Aspects of the National
Flood Insurance Program (GAO/T-RCED-93-70, Sept.  14, 1993). 


   RESULTS IN BRIEF
------------------------------------------------------------ Letter :1

The flood insurance program is intentionally not actuarially sound
because the Congress authorized subsidized insurance rates to be made
available for policies covering certain structures.  Because about 41
percent of policies were subsidized as of 1993, overall premium
income, while sufficient to cover flood losses sustained in most
recent years, is not sufficient to build reserves to meet future
expected flood losses.  The Insurance Administration's annual target
for the program's overall premium income is the amount of loss in an
average historical loss year, which is the approximate average annual
loss experience under the program since 1978.  Since no catastrophic
loss years have occurred since 1978, collecting premiums that are
based on an average historical loss year does not enable the fund to
build sufficient reserves to cover a possible catastrophic loss year
in the future.  Thus, it is inevitable that claims losses and program
expenses will exceed the program's funds in some years. 

Increasing the premiums charged to subsidized policyholders (thereby
decreasing the subsidy) to improve the program's financial health
could have an adverse impact on other federal disaster-related relief
costs.  Increasing subsidized rates would be likely to cause some
policyholders to cancel their flood insurance, and if flooded in the
future, these people might apply for Small Business Administration
loans or FEMA disaster assistance grants.  On the other hand, efforts
to build reserves by increasing participation in the flood insurance
program would be likely to reduce the costs of other disaster
assistance programs, but these efforts could also worsen the flood
insurance program's financial condition by increasing the number of
subsidized policyholders in the program. 

The Insurance Administration sets rates for post-FIRM construction on
the basis of actuarial principles that consider the actual flood risk
of an insured structure, such as whether a structure is inside a
special flood hazard area.  The policies with these rates are not
subsidized by the federal government.  For structures covered by
subsidized rates, which include flood-prone pre-FIRM structures, the
Insurance Administration sets subsidized rates to generate sufficient
premium income so that overall program premiums from both actuarial
and subsidized policies approximate the amount of an average
historical loss year.  Despite subsidized premiums, the Insurance
Administration expects the average premium for a subsidized policy to
be about $401 in 1994 and the average premium for an actuarial policy
to be about $247.  The higher average premium for a subsidized policy
reflects the significantly greater riskiness of flood-prone pre-FIRM
properties. 


   BACKGROUND
------------------------------------------------------------ Letter :2

Over 18,000 communities have joined the flood insurance program.  The
FIRMs prepared for the Insurance Administration by the U.S.  Army
Corp of Engineers and private engineering companies for these
communities identified special flood hazard areas, also known as the
100-year floodplains, which are areas subject to a 1-percent or
greater chance of experiencing flooding in a given year.  A key
component of the program's building standards that must be followed
by communities participating in the program is a requirement that the
lowest floor of the structure be elevated to or above the base flood
level--the elevation at which there is a 1-percent chance of flooding
in a given year. 

To encourage communities to join the program, thereby promoting
floodplain management and widespread purchasing of flood insurance,
the Congress authorized the Insurance Administration to make
subsidized flood insurance rates available to owners of structures
built before a community's FIRM was prepared.  These pre-FIRM
structures are generally more flood-prone than later-built structures
because they were not built according to the program's building
standards.  However, owners of pre-FIRM properties that are
sufficiently elevated can opt for actuarial rates. 

From 1968 until the adoption of the Flood Disaster Protection Act of
1973, the purchase of flood insurance was voluntary.  The 1973 act
required the mandatory purchase of flood insurance to cover
structures in special flood hazard areas of communities participating
in the program if (1) any federal loans or grants were used to
acquire or build the structures and (2) loans were secured by
improved properties and the loans were made by lending institutions
regulated by the federal government.  Owners of properties with no
mortgages or properties with mortgages held by unregulated lenders
are not required to buy flood insurance, even if the properties are
in special flood hazard areas. 

For the program to be actuarially sound, its rate-setting process
would have to include a consideration of the monetary risk exposure
of the program, or the dollar value of expected flood losses over the
long run.  Since the magnitude of flood damage varies considerably
from year to year, premium income in many years would exceed actual
losses.  This circumstance would enable the fund to build reserves
toward a possible catastrophic year in the future. 

Supplementing the program, the Small Business Administration offers
low-interest loans to flood victims who are creditworthy.  A flood
victim who cannot obtain a Small Business Administration loan may
apply for an individual and family FEMA grant of up to $11,900 or the
amount of the loss, whichever is less. 


   THE PROGRAM IS NOT, NOR WAS IT
   INTENDED TO BE, ACTUARIALLY
   SOUND
------------------------------------------------------------ Letter :3

The program is not actuarially sound by intention.  The Congress
authorized the Insurance Administration to subsidize a significant
portion of the total policies in force, although it did not provide
annual appropriations to cover the implicit subsidy. 

The Congress also authorized the Insurance Administration to borrow
up to $1 billion from the U.S.  Treasury if necessary to pay claims
losses.  Also, since the inception of the program in 1968 through
fiscal year 1986, the Congress appropriated about $2.1 billion (which
represents about $3.3 billion in constant 1992 dollars) to the
program; about half of the appropriation was to repay past loans from
the U.S.  Treasury, and the other half was to pay for administrative
expenses.  However, no appropriations have been made to the program
since fiscal year 1986. 


      SUBSIDIZED RATES RESTRICT
      THE PROGRAM'S INCOME AND
      COVER STRUCTURES THAT INCUR
      GREATER FLOOD DAMAGE
---------------------------------------------------------- Letter :3.1

The program is not actuarially sound because about 41 percent of the
2.7 million policies in force are subsidized.  For a single-family
pre-FIRM property, subsidized rates are available for the first
$35,000 of coverage, although any insurance coverage above that
amount must be purchased at actuarial rates.  The Insurance
Administration computed that total premiums paid by subsidized
policyholders in fiscal year 1991 were about $780 million less than
if these rates had been actuarially based and participation had
remained the same.  While the Insurance Administration only estimated
the dollar value of the subsidy for this one year, the fund would
currently have a significant reserve if rates had never been
subsidized and participation in the program had not been affected by
higher rates. 

Pre-FIRM structures that are within an identified 100-year floodplain
and are covered by subsidized policies are, on average, not as
elevated as the post-FIRM structures in comparison with the base
flood level.  Insurance Administration officials told us that, on
average, pre-FIRM structures not built to the program's standards are
4-1/2 times more likely to suffer a flood loss.  When these
structures suffer a loss, the damage sustained is, on average, about
one-third greater than the damage to flooded post-FIRM structures. 
According to the Insurance Administration, when these two factors are
combined, pre-FIRM structures suffer, on average, about 6 times more
damage than post-FIRM structures. 


      PREMIUM INCOME NOT
      SUFFICIENT TO BUILD RESERVES
      FOR POTENTIAL CATASTROPHIC
      LOSSES
---------------------------------------------------------- Letter :3.2

As an alternative to actuarial soundness, the Insurance
Administration developed a financial goal for the program to collect
sufficient revenues to at least meet the expected losses of the
average historical loss year, as well as to cover all
non-loss-related program expenses, such as program administration. 
However, the average historical loss year is based only on the
experience under the program since 1978.  Since that time, no
catastrophic year has occurred,\3 and many years in the 1980s were
characterized by fairly low loss levels.  Therefore, the average
historical loss year involves less claims losses than the expected
per annum claims losses in future years, and collecting premiums to
meet the average historical loss year does not reflect the
collections necessary to build reserves for potential catastrophic
years in the future. 

The Insurance Administration determines the overall revenue
requirements necessary to meet an average historical loss year
through an analysis of a variety of reports about previous years'
policies and claims.  Since the numbers and types of policies can
change from year to year, past experience is used to determine how
such changes should be accounted for when determining future revenue
needs.  Additionally, the average historical loss year must be
adjusted for inflation, since a given amount of actual damage to a
structure in a previous year will have generated a smaller dollar
claim than the same damage will generate in a current or future year. 
Finally, any changes in coverage offered under the program's policies
would need to be considered when revenue needs are determined. 

The level of the average historical loss year will change over time
because of inflation, changes in the number and types of policies,
and changes in loss levels.  For example, while the average
historical loss year was about $390 million in fiscal year 1993, it
rose to about $450 million for 1994 because of a recent increase in
the number of policies, the high loss experience in fiscal year 1993,
and inflation. 

Because rates for actuarial policies include a catastrophic risk
provision, while subsidized rates do not, the contribution made
toward the average historical loss year, as well as the "long-run
expected loss year," differs considerably for actuarial and
subsidized policies.  For example, by design, premium income from
actuarial policies in the 100-year floodplain equals 100 percent of
those policies' expected claims losses over the long run, but the
same premium income accounts for 124 percent of the expected claims
losses of an average historical loss year.  Conversely, premiums from
subsidized policies account for only 33 percent of the premiums
necessary to pay the expected long-run claims losses on these
policies and 92 percent of an average historical loss year for those
policies.  The drain on the program comes from these policies. 
Proposed rate increases in fiscal year 1994 would raise subsidized
contributions to 97 percent of the average historical loss year. 


--------------------
\3 Insurance Administration officials told us that a catastrophic
year resulting in $3 billion to $4 billion in claims losses has a 1
in 1,000 chance of occurring. 


      SEVERE RECENT FLOODING
      RESULTED IN THE INSURANCE
      ADMINISTRATION'S EXERCISING
      ITS BORROWING AUTHORITY
---------------------------------------------------------- Letter :3.3

Between fiscal year 1987 and the end of fiscal year 1993, the
Insurance Administration's goal for the program of basing revenues on
the amount of the average historical loss year, instead of on a
long-run expected loss year (about twice as much), has allowed the
fund to cover insurance claims as well as program and administrative
costs without borrowing from the U.S.  Treasury.  However, the nation
experienced severe flood damage in fiscal year 1993, primarily
because of the December 1992 nor'easter, the March 1993 flooding in
western Florida, and the July 1993 Midwest flooding.  Claims during
fiscal year 1993 were about $984 million, which is more than double
the average historical loss year. 

At the end of fiscal year 1993, according to Insurance Administration
officials, the fund had a positive cash balance, but the fund's
obligations for outstanding claims were about $110 million more than
the program's available resources.\4 In December 1993, the Insurance
Administration borrowed $100 million from the U.S.  Treasury.  As of
February 1994, about $12 million has been used to pay claims. 
Whether the Insurance Administration will have to further exercise
its borrowing authority will depend on (1) the relative timing of
payments on its current obligations and expected monthly premium
receipts of about $55 million and (2) future insurance claims. 


--------------------
\4 Insurance Administration officials noted that beginning in fiscal
year 1986, the Congress required all program and administrative costs
to be paid for by the fund without a commensurate rate increase.  In
1991, the Congress authorized the Insurance Administration to charge
policyholders a federal policy fee to pay for these costs.  However,
because costs were not collected between 1986 and 1991, program
assets were reduced by about $355 million, according to Insurance
Administration officials.  These officials noted that the fund at the
end of 1993 would have had a positive balance of $245 million had
these costs been funded from commensurate premium increases. 


   INCREASING PREMIUMS FOR
   SUBSIDIZED POLICIES OR
   EXPANDING PARTICIPATION IN THE
   PROGRAM MAY HAVE ADVERSE
   FINANCIAL IMPACTS
------------------------------------------------------------ Letter :4

The government's expenditures on disaster assistance include both
direct disaster payments through Small Business Administration loans
and FEMA grants as well as the government's costs of subsidizing some
flood insurance rates.  The amount of direct disaster assistance is
the result, in part, of the level of participation in the flood
insurance program. 

Therefore, efforts to build reserves to improve the financial health
of the program by charging higher premiums for subsidized
policyholders may not minimize the federal government's overall
expenditures on flood-related disaster relief.  On the other hand,
expanding participation in the program is likely to reduce the cost
of other federal efforts to provide flood relief, but greater
participation by subsidized property owners could increase the
program's unfunded liability. 

Two bills introduced in the 103rd Congress--S.  1405 and H.R. 
62--would revise the program by, among other things, expanding
participation in the program by increasing compliance with the
mandatory purchase requirement or extending the mandatory purchase
requirement to mortgages not held by federally regulated lenders. 
Both bills would establish an interagency task force to conduct
studies and make recommendations to revise the program.  S.  1405
requires the task force to study the possibility of revising the rate
structure to account for catastrophic events and propose strategies
to establish an actuarially based premium structure to account for
all insurable risks. 


      INCREASING PREMIUMS ON
      SUBSIDIZED POLICIES MAY LEAD
      SOME POLICYHOLDERS TO CANCEL
      THEIR POLICIES
---------------------------------------------------------- Letter :4.1

Increasing premiums on subsidized policies may not minimize the
federal government's overall expenditures on flood-related disaster
relief.  Because they were built before the program's building
standards became applicable, pre-FIRM structures are generally not as
elevated as post-FIRM structures, and if their owners were to be
charged true actuarial rates, these rates would be much higher than
current subsidized rates.\5 For example, if the subsidy on pre-FIRM
structures were eliminated, insurance rates on currently subsidized
policies would need to rise, on average, approximately threefold,
implying an annual average premium of about $1,100 for these
structures.  Significant rate increases for subsidized policies,
including charging actuarial rates, would be likely to cause some
pre-FIRM property owners--we do not know how many--to cancel their
flood insurance.  Although the information is dated, our analysis in
the early 1980s indicated that if the program doubled the
then-existing average premiums (both subsidized and actuarial), about
40 percent of the policyholders would be expected to cancel their
policies.\6

If owners of pre-FIRM structures, which suffer the greatest flood
loss, canceled their insurance policies, the federal government would
be likely to face increased costs, as the result of future floods, in
the form of low-interest loans from the Small Business Administration
or grants from FEMA.  The effect on total federal disaster assistance
costs of phasing out subsidized rates depends on the number of the
program's current policyholders who would cancel their policies,
which is unknown.  Thus, it is difficult to estimate if the increased
costs of other federal disaster relief programs would be less than or
more than the cost of the program's current subsidy. 


--------------------
\5 Also, Insurance Administration officials told us that making all
rates actuarially based would not make the program actuarially sound. 
They noted that an initial capitalization would be necessary to
establish some reserves in the event that a catastrophic year were to
occur before sufficient reserves were accumulated from premium
income. 

\6 National Flood Insurance Program:  Major Changes Needed If It Is
to Operate Without a Federal Subsidy (GAO/RCED-83-53, Jan.  3, 1983). 


      EXPANDING PARTICIPATION WILL
      INCREASE THE PROGRAM'S
      POTENTIAL LIABILITY
---------------------------------------------------------- Letter :4.2

Expanding participation in the program by increasing the rate of
compliance with the mandatory purchase requirement or by extending
the mandatory purchase requirement to property owners not now covered
will be likely to increase the number of both subsidized and
unsubsidized policies.  Although greater participation in the program
is likely to reduce the cost of FEMA grants and Small Business
Administration loans, the resultant increase in subsidized
policyholders will put greater financial stress on the flood
insurance program because the premiums received from subsidized
policyholders are not sufficient to meet the future estimated losses
on these policies. 


   THE PROGRAM'S RATE-SETTING
   PROCEDURES
------------------------------------------------------------ Letter :5

Insurance rate-setting for national flood insurance differs,
depending on whether a structure is covered by actuarial or by
subsidized insurance rates.  Subsidized insurance rates are available
for pre-FIRM structures.  However, post-FIRM structures, and certain
pre-FIRM structures that qualify, are assessed actuarial rates. 

Flood insurance can cover a structure and/or its contents.  The
maximum limits differ, depending on the structure; for example, the
limits for a single-family structure are $185,000 for the structure
and $60,000 for its contents. 


      ACTUARIAL RATES ARE BASED ON
      ACTUAL RISK EXPOSURES
---------------------------------------------------------- Letter :5.1

Rates for post-FIRM construction are actuarial and are not subsidized
by the federal government.  The Insurance Administration's method for
establishing these rates for post-FIRM structures lying within the
100-year floodplain follows a hydrologic method that is based on
studies performed by the U.S.  Army Corps of Engineers and private
engineering companies.  These rates are based on available hydrologic
data, flood insurance claims, and simulations, as well as on
engineering and actuarial judgment.\7 According to the Insurance
Administration, the basic data elements it needs to predict expected
flood loss include (1) probability estimates of the frequency with
which floods of different severity will occur and (2) estimates of
structural property damage caused by different types of floods.  The
Insurance Administration accounts for several program expense items,
such as agents' commissions and the program's administrative costs,
in the actuarial rates.  (See app.  II for more details on the
actuarial rate-setting process.)

Actuarial rates are based on actual risk exposures and generally vary
according to several risk-related factors.  The following are the
most important of these factors: 

  The flood-risk zone.  Owners of structures in zones subject to
     greater flooding risk pay higher rates than owners of structures
     in zones that have less severe flood risk. 

  The elevation of the structure relative to the base flood level. 
     Even within a given flood-risk zone, the higher a structure is
     elevated relative to the base flood level, the lower the rates
     charged, because buildings at a higher elevation face a lower
     risk of flooding. 

  The amount of insurance purchased.  Rates vary depending on how
     much insurance is being purchased.  The Insurance Administration
     sets rates for the "first layer" (the first $45,000 of insurance
     purchased on a single-family dwelling) at a higher rate than for
     coverage above that amount.  This feature of the program's rate
     structure reflects differential risks, since claims are more
     likely to be made against the first several thousand dollars of
     coverage than against much higher levels of coverage. 


--------------------
\7 We have not independently reviewed the studies on which the
Insurance Administration's data for actuarial rate-setting are based. 


      SUBSIDIZED RATES ARE SET BY
      ADMINISTRATIVE AND
      LEGISLATIVE PROCEDURES
---------------------------------------------------------- Letter :5.2

Owners of buildings constructed before the completion of a
community's FIRM or before January 1, 1975, whichever is later, can
purchase subsidized insurance.  In 1993, about 41 percent of the
program's policies were subsidized, but this percentage will decline
over time as newer properties join the program and are charged
actuarial rates.  Subsidized rates on pre-FIRM properties have never
been set by an analysis of the underlying flood risk.  Instead, they
are set by an administrative and legislative process.  Insurance
Administration officials stated that the use of the average
historical loss year as an overall financial goal for the program
helps to provide a more objective standard for the setting of
subsidized rates than was true in the past. 

To encourage greater participation in the program, rates for
subsidized policies were decreased during the 1970s.  By contrast, in
the 1980s subsidized rates were raised, and coverage became more
limited as the Insurance Administration attempted to meet its
financial goal of collecting revenues sufficient to at least meet an
average historical loss year.  Insurance Administration officials
said that they would keep taking steps to make subsidized rates more
reflective of their actual risk exposure by decreasing policy
coverage and increasing policy deductibles. 

For setting rates on subsidized policies, the Insurance
Administration's current method is to first determine the revenue
needed to cover non-loss-related costs, such as that for program
administration, as well as to collect sufficient premiums to at least
meet an average historical loss year, on the basis of the current
policies in force and the current price level.  Next, the Insurance
Administration determines the revenue it will receive from policies
with actuarially based rates.  The Insurance Administration then
subtracts the expected revenue from actuarially based policies from
the average historical loss year level to determine the minimum
premium income needed from policies with subsidized rates.  Finally,
the Insurance Administration computes the subsidized rates on the
basis of the minimum revenue needed and the expected number of
subsidized policies.  The proposed subsidized rates are published in
the Federal Register for public comment and submitted for
congressional approval as part of the Insurance Administration's
budget and authorization proceeding. 

For single-family pre-FIRM properties, subsidized rates are available
only on the first $35,000 of insurance coverage; rates for any
additional insurance coverage are actuarially based.  Although
subsidized, rates for the first $35,000 of coverage for single-family
pre-FIRM properties are generally significantly higher than actuarial
rates for the first $35,000 of coverage on single-family post-FIRM
structures that were built in compliance with the program's building
standards.  For example, the actuarial rate on the first layer of
coverage for a one-story single-family post-FIRM structure with no
basement in an AE zone (an Insurance Administration-designated flood
zone lying within the 100-year floodplain) that is built at the
elevation of the base flood level is 33 cents per $100 of insured
value.  On the other hand, the subsidized rate on a similarly located
pre-FIRM structure that does not meet the program's building
standards is 55 cents per $100 of insured value.  Subsidized rates
are generally higher than actuarial rates because of the
substantially greater flood risk posed by pre-FIRM properties when
they are compared to well-situated post-FIRM properties. 

Because of the lower rates for actuarially based policies, owners of
pre-FIRM construction often apply for actuarial rates if they can
qualify to do so.  In order for a pre-FIRM structure lying within the
100-year floodplain to qualify for actuarial rates, the owner must
obtain an elevation certificate that specifies that the lowest floor
of the structure is at least at the base flood level.  Approximately
29 percent of all of the program's policies are for pre-FIRM
structures that have qualified for actuarial rates. 


   CONCLUSIONS
------------------------------------------------------------ Letter :6

The program is not actuarially sound because it does not collect
sufficient premium income to build reserves to meet future expected
flood losses.  This situation occurs because premiums for 41 percent
of the program's policies are subsidized.  Therefore, it is
inevitable that claims losses and program expenses will exceed the
funds available to the program in some years, and if a catastrophic
loss year were to occur, not even the Insurance Administration's
borrowing authority would be sufficient to cover claims losses. 

Efforts have been made recently, such as S.  1405 and H.R.  62, to
study revising the program by increasing the premiums paid by
subsidized policyholders and expanding program participation. 
Increasing the premiums paid by subsidized policyholders to the
actuarial level, or to some level between the current rate and the
actuarial rate, may improve the program's financial health.  However,
higher premiums would also be likely to increase the costs of other
disaster-related relief programs, because some policyholders would
cancel their insurance but would receive other disaster assistance
grants or loans, in the event of a flood.  On the other hand,
increased participation in the program by subsidized and unsubsidized
property owners is likely to reduce the cost of other federal
disaster assistance programs, but greater participation would also
put greater financial stress on the program, to the extent that
additional subsidized properties are covered under the program. 


   MATTERS FOR CONGRESSIONAL
   CONSIDERATION
------------------------------------------------------------ Letter :7

Any attempt by the Congress to revise the flood insurance program in
ways that will affect program participation, such as by expanding or
strengthening the mandatory purchase requirement, should be
considered in the context of the integral relationship between this
program and other disaster assistance programs.  Similarly, a
revision of the subsidized premium rate structure that would
eliminate all or part of the present subsidy should be analyzed in
the context of the potential financial impact on other federal
disaster assistance programs through, for example, the possible
cancellation of policies by policyholders. 


   AGENCY COMMENTS
------------------------------------------------------------ Letter :8

We discussed the facts of this report with the Federal Insurance
Administration's Deputy Administrator, the Executive Assistant to the
Federal Insurance Administrator, and FEMA's Acting Deputy Chief
Financial Officer.  They generally agreed with our facts as presented
on the actuarial soundness of the program, the potential financial
impacts of either increasing subsidized flood insurance rates or
expanding program participation, and the program's rate-setting
procedures.  They also agreed with our matters for congressional
consideration.  We incorporated, where appropriate, changes suggested
by the officials to clarify certain information presented.  As
requested, we did not obtain written agency comments on a draft of
this report. 


   SCOPE AND METHODOLOGY
------------------------------------------------------------ Letter :9

We reviewed the actuarial soundness of the program and rate-setting
procedures for both actuarial and subsidized policies.  To complete
this work, we reviewed the Insurance Administration's documents and
interviewed relevant officials, including the Deputy Administrator of
the Federal Insurance Administration and an Executive Assistant to
the Administrator.  In addition, we reviewed literature on actuarial
rate-setting and spoke with two officials--one from a state insurance
agency and one from a private insurance organization--familiar with
the flood insurance program. 

To identify the potential financial impacts of increasing subsidized
rates and expanding program participation, we held discussions with
Insurance Administration officials, analyzed data obtained from our
review work mentioned above, and analyzed proposed legislation. 

In reviewing the financial management of the program, we relied on
the FEMA Inspector General's recently completed reviews of the
program's financial statements.  We talked to members of the
Inspector General's staff and examined their workpapers.  We also
interviewed FEMA's Chief Financial Officer and flood insurance
program staff responsible for the financial management of the
program. 

We conducted our review from January through September 1993 and
updated certain information through February 1994. 


---------------------------------------------------------- Letter :9.1

Unless you announce its contents earlier, we plan no further
distribution of this report until 30 days from the date of this
letter.  At that time, we will send copies to interested
congressional committees; the Director, FEMA; the Director, Office of
Management and Budget; and other interested parties.  We will make
copies available to others upon request. 

This work was performed under the direction of Judy A. 
England-Joseph, Director of Housing and Community Development Issues. 
If you or your staff have any questions, she can be reached at (202)
512-7632.  Major contributors to this report are listed in appendix
III. 

Keith O.  Fultz
Assistant Comptroller General


ACTIONS TAKEN BY FEMA ON THE
NATIONAL FLOOD INSURANCE PROGRAM'S
FINANCIAL MANAGEMENT PROBLEMS
=========================================================== Appendix I

Since 1979, FEMA has acknowledged problems with its financial
operations.  In 1989, the Office of Management and Budget (OMB)
designated FEMA's financial management system and internal controls
program as high-risk areas.  In January 1993, OMB voiced reservations
about the adequacy of FEMA's progress and future corrective action
plans.  Some of FEMA's financial management problems affect the flood
insurance program. 

FEMA's Office of Inspector General (OIG) audited FEMA's financial
statements for fiscal years 1991 and 1992 in accordance with the
Chief Financial Officer's Act.  The audits identified problems in the
program's financial management system and internal control structure
that prevented accumulation and reporting of reliable financial
information.  That information includes the fund's balance on deposit
in the U.S.  Treasury and its other assets and costs of operation. 

On September 14, 1993, we testified on these issues before the
Subcommittee on Housing and Urban Affairs, Senate Committee on
Banking, Housing, and Urban Affairs.  Subsequent to our testimony,
FEMA took action, or agreed to take action, to correct the problems
identified in the OIG audits. 


   FEMA'S RECORDS DID NOT
   EFFECTIVELY TRACK OR MONITOR
   THE PROGRAM'S FUND BALANCE
--------------------------------------------------------- Appendix I:1


      OIG'S AUDIT RESULTS
------------------------------------------------------- Appendix I:1.1

As a result of a review of the program's financial statements as of
September 30, 1991, FEMA's Inspector General concluded that FEMA does
not have systems or records to effectively track or monitor the
program's fund balance with the U.S.  Treasury.  He also reported
that the program's cash balance with the U.S.  Treasury is commingled
with all other FEMA funds.  Furthermore, for many years FEMA has not
consistently reconciled its records with reported U.S.  Treasury
funds.  Therefore, FEMA's Inspector General reported that it could
not verify the fund balance with the U.S.  Treasury.  The Inspector
General recommended that FEMA begin reconciling the fund's cash
balance on deposit at the U.S.  Treasury each month. 

Although FEMA reconciled the balance for only some months during
fiscal year 1992, the Inspector General was able to validate the
program's receipts and disbursements for fiscal year 1992. 
Subsequently, after reviewing the program's statements as of
September 30, 1992, the Inspector General again reported that it
could not verify the reported ending fund balance because the
beginning balance could not be verified and no separate account
existed at the U.S.  Treasury for the fund.  The Inspector General
recommended that FEMA establish a separate account in the U.S. 
Treasury for the flood insurance fund. 


      FEMA'S RESPONSE
------------------------------------------------------- Appendix I:1.2

FEMA officials acknowledged the inaccuracies with the fund balance
and stated that they were working at resolving the problems and
anticipate performing reconciliations needed to validate the account
balance.  They stated that the potential misstatements concerning the
fund's balance were not significant enough to affect the day-to-day
decisions they make to ensure that sufficient resources are available
to continue program operations and to make payments on claims. 


      GAO'S TESTIMONY
------------------------------------------------------- Appendix I:1.3

We concluded in our testimony that FEMA's determination of when, and
if, the flood insurance fund needs to borrow from the U.S.  Treasury
may not be based on adequate data that FEMA maintains on the amount
of program funds it has on deposit in the U.S.  Treasury.  Because
FEMA may have to exercise its borrowing authority to pay claims, we
stated that FEMA should implement the OIG's recommendation to
establish a separate flood insurance program balance in the U.S. 
Treasury. 


      ACTION TAKEN BY FEMA
------------------------------------------------------- Appendix I:1.4

Subsequent to our testimony, in October 1993 FEMA received
notification from the Financial Management Service of the Department
of the Treasury that confirms the establishment of a separate fund
balance at the U.S.  Treasury for the flood insurance fund. 


   FEMA'S FINANCIAL REPORTING IS
   UNRELIABLE
--------------------------------------------------------- Appendix I:2


      OIG'S AUDIT RESULTS
------------------------------------------------------- Appendix I:2.1

FEMA's Inspector General was unable to express an opinion on the
program's financial statements for fiscal years 1991 and 1992 because
it found that (1) property and equipment accountability was
inadequate, (2) inventories of its flood maps were not accounted for,
and (3) administrative expenses were not accurately reported. 

In the area of property and equipment accountability, FEMA's
Inspector General reported that FEMA's policy is to expense property
and equipment, such as digital and engineering equipment, as
acquired.  FEMA has no system to monitor or track its property,
including its property and equipment held by contractors. 

A contractor maintains FEMA's inventory of flood maps in a warehouse
and purportedly has a perpetual inventory system to account for the
quantities.  However, FEMA did not report any flood insurance program
inventory in its fiscal year 1991 financial statements, and in fiscal
year 1992 it reported an inventory amount based on estimated
quantities and unit costs rather than on a physical inventory.  The
costs of the maps, estimated to be 15 cents per map, are not
documented, and as a result, the Inspector General could not validate
the reported inventory of $6 million. 

FEMA policy requires administrative expenses to be reported on an
accrual basis that ensures that revenues and expenses are matched to
the period in which the revenue is earned or the costs are incurred. 
However, the Inspector General found that some expenses incurred in
administering the program are reported on an obligation basis.  As a
result of the inconsistent reporting, FEMA's Inspector General
concluded that the financial reports do not provide an accurate,
reliable perspective on the costs incurred nor the results of
operations for the fiscal year. 

The Inspector General made various recommendations, including
short-term actions to correct these problems.  For example, FEMA
should conduct physical inventories of the program's assets and
report administrative expenses on an accrual basis. 


      FEMA'S RESPONSE
------------------------------------------------------- Appendix I:2.2

FEMA agreed to take some steps to improve its financial management
problems, but it did not agree to implement either of the above
recommendations.  Instead, it said it would rely on the
implementation of FEMA's Five-Year Financial Plan for fiscal years
1992-96 to provide long-term solutions to the program's financial
management problems. 


      GAO'S TESTIMONY
------------------------------------------------------- Appendix I:2.3

We concluded in our testimony that FEMA should reexamine its decision
not to make short-term improvements in the program's financial
management system. 


      ACTIONS TAKEN BY FEMA
------------------------------------------------------- Appendix I:2.4

FEMA officials told us that, as discussed in our testimony, they have
decided to make improvements in the program's financial management
system.  FEMA officials said they are in the process of awarding a
new contract to maintain their inventory of flood maps.  The new
contract will require the contractor annually to make a physical
inventory of the maps.  FEMA officials told us that, other than the
flood maps, they have very little equipment or property at
contractors.  The physical inventory required by the new contractor
will correct the problems noted in the OIG report concerning property
accountability, according to these officials.  Also, FEMA officials
told us that they are now reporting expenses on an accrual basis as
recommended in the OIG report. 


ACTUARIAL RATE-SETTING FOR
POST-FIRM STRUCTURES
========================================================== Appendix II

This appendix discusses the methodology used by the Federal Insurance
Administration (FIA) for setting actuarial rates for structures that
were built after flood insurance rate maps (FIRM) were prepared
(referred to as post-FIRM construction).  As discussed in the letter
of this report, actuarial rates are charged on post-FIRM construction
and on pre-FIRM construction that have been certified as meeting the
National Flood Insurance Program's elevation standards.  This
appendix (1) describes actuarial rate-setting, (2) discusses key
characteristics used to classify post-FIRM properties according to
flood risk, (3) describes data elements necessary and the methodology
used in the application of the hydrologic model for actuarial
rate-setting, (4) discusses other components of actuarial rates, and
(5) provides examples of post-FIRM actuarial rates for properties
lying both inside and outside of the 100-year floodplain. 


   ACTUARIAL RATES ARE BASED ON
   FLOOD RISK
-------------------------------------------------------- Appendix II:1

Insurance is a mechanism through which policyholders can pay a
specific price in order to transfer a risk that they face to some
other entity.  For example, homeowners face the risk that their
houses will burn down.  By purchasing insurance at a predetermined
price, a homeowner can effectively transfer, to an insurance company,
most of the financial risk associated with losing a house to fire. 

For a private-sector insurance firm to offer such risk transfer
through the provision of insurance and remain profitable, it must set
insurance rates high enough to cover expected claims losses, as well
as non-loss-related expenses.  To do this, a firm needs to set
insurance rates in accordance with risk exposure or, in other words,
the expected financial loss that the firm takes on by providing
insurance on the current set of policies in force.  Insurance rates
that are set by taking into consideration estimated risk exposure are
known as actuarial rates. 

In the case of flood insurance, FIA uses a class-rating rather than
an individual-rating system.  That is, FIA classifies properties
according to key characteristics of flood risk.  All owners of
properties in the same group are then charged the same rates.  Even
though individual risks may vary among properties within each risk
group classification, these rates are actuarial in the sense that
risk exposure for each classification of like properties is taken
into consideration when setting the group's rates. 


   KEY CHARACTERISTICS USED TO
   CLASSIFY PROPERTIES ACCORDING
   TO FLOOD RISK
-------------------------------------------------------- Appendix II:2

In order to set actuarial rates for national flood insurance,
information about the risk of flooding is essential.  One of the
primary objectives of the National Flood Insurance Act was to
identify flood-prone areas.  In doing so, flood insurance rate maps,
which have been completed for nearly all communities that were
considered to be flood-prone, provide information that is crucial for
classifying properties according to flood risk.  The key
characteristics that are used to classify properties according to
flood risk include the flood zone and the elevation of a structure
relative to the base flood elevation (BFE).  Information about both
the zone and the BFE are obtained from FIRMs. 


      FLOOD ZONES
------------------------------------------------------ Appendix II:2.1

Knowledge of the flooding risk zone is important for actuarial
rate-setting because areas of differential flood risk should be
charged different rates.  Each FIRM outlines the flood zones
throughout the community.  The zones with a first letter of either A
or V, are classified as "special flood-hazard areas" (SFHA).  These
areas are believed to face a 1-percent or greater chance of being
flooded in a given year and are also known as the 100-year
floodplain.  V zones include coastal areas that incur wind velocities
and associated wave heights that pose additional risks to properties
during flooding events.\1 The other major zone is zone X, which
includes areas outside the identified 100-year floodplain. 


--------------------
\1 The majority of policies under the flood insurance program are for
structures in an A zone or the X zone.  Relatively few structures are
in the V zones, accounting for less than 2 percent of all post-FIRM
properties. 


      BASE FLOOD ELEVATION
------------------------------------------------------ Appendix II:2.2

The maps, in most cases, also delineate the BFE for areas that lie
within the identified 100-year floodplain.  The BFE is the elevation
relative to mean sea level at which there is a 1-percent chance of
flood waters rising to 1 foot or more in a given year.  The level of
the BFE within a community can change throughout the floodplain, and
those changes are delineated on FIRMs.  The establishment of the BFE
in V zones also takes into account the elevation of storm surges and
the expected height of wave crests above storm surges.  Thus, they
are called BFEWH--"base flood elevation with wave height."

Knowledge of the BFE is important for a couple of reasons.  First,
the program's building standards require that the bottom level of
structures be built at least to the elevation of the BFE to ensure
that structures are not subject to a greater than 1-percent chance of
flooding in a year.  Second, in terms of classifying properties
according to flood risk, knowledge of the BFE is important, because
flood risks vary with the elevation of a structure relative to the
BFE:  The more elevated a structure is, the less likely flood waters
will reach it.  Thus, rates are set so that structures with the
lowest floor elevated above the BFE are charged lower rates than
those elevated only to the BFE or below the BFE.\2


--------------------
\2 Although post-FIRM properties are supposed to be elevated at least
to the BFE, some properties may not meet this code.  FIA provides
rates for properties out of compliance on rate sheets as long as they
are not more than 1 foot below the BFE.  Rates for properties with
the lowest floor elevated below the BFE are considerably higher.  If
a post-FIRM structure is more than 1 foot below the BFE, rates can be
obtained by submitting to FIA directly. 


   THE HYDROLOGIC MODEL\3
-------------------------------------------------------- Appendix II:3

The basic method for establishing actuarial rates on post-FIRM
construction lying within the 100-year floodplain follows the
hydrologic model described in a 1966 report by the Department of
Housing and Urban Development entitled Insurance and Other Programs
for Financial Assistance to Flood Victims.\4 The basic logic of the
hydrologic model is to set flood insurance rates for a property
according to its risk of being flooded.  Thus, a major portion of
flood insurance rates is based on the per annum expected dollar flood
loss for a property of a given classification. 

In the previous section, we noted that the key characteristics of the
zone and the elevation of a structure relative to the BFE, which are
available from FIRMs, are important for categorizing post-FIRM
properties according to flood risk.  Certain characteristics of the
property, such as whether it has a basement, are also used for
classifying properties according to risk.  However, once a property
is thus classified, all properties within the same group are charged
the same flood insurance rates. 

In the remainder of this section, we discuss the primary data sources
necessary to apply the hydrologic model of rate-setting.  These data
provide information on the risk of flood for a given type of property
within a given zone and of a particular relative elevation.  We also
discuss the hydrologic rate-setting model. 


--------------------
\3 The method discussed in this section is used to determine
actuarial rates for both the building structure as well as the
insured's personal belongings, or the "contents" contained within the
structure.  Rates differ for these types of coverage, and buyers
specifically purchase each type of coverage in order to be covered
for both.  Most of this discussion, however, focuses on coverage for
the structure. 

\4 Rates for post-FIRM properties in zone X, which are outside the
100-year floodplain, are set primarily through an analysis of
previous years' claims. 


      ELEVATION-FREQUENCY
      RELATIONSHIP--"PELV" VALUES
------------------------------------------------------ Appendix II:3.1

A very important data element needed for the application of the
hydrologic model is an estimate of the probabilities that floods of
different severities, relative to the BFE, will occur in a given
year.  FIA calls these data probability of elevation (PELV) values. 
Within any zone, there is a 1-percent chance that flood waters will
reach the BFE.  However, across zones the likelihood that flood
waters will reach 1-foot above or below that level will vary.  For
example, FIA notes that in zone A10 (currently part of zone AE), the
probability of water rising to or above 1 foot below the BFE is 1.6
percent per year, and the probability of water reaching or exceeding
1 foot above the BFE is 0.6 percent per year. 

PELV tables provide detailed information, by zone, about the
frequency with which we can expect floods of all possible water
surface elevations to occur.  These data were generated on the basis
of detailed engineering studies, available flood insurance data,
simulations, and professional judgments and were established for each
flood-hazard zone to meet generally accepted scientific parameters
and legal considerations.  One of the problems in establishing PELV
tables, however, was that the flood histories on which these studies
were based were generally not very long.  Statistical literature has
shown that when the history of these events is too short, the number
of occurrences is generally, small which causes a bias toward
establishing frequency probabilities that are too low.  Consequently,
the original PELV values were modified to account for this
statistical bias. 


      DEPTH-DAMAGE
      RELATIONSHIP--"DELV" VALUES
------------------------------------------------------ Appendix II:3.2

A second necessary data element for the hydrologic model are
estimates of the structural damage that will be suffered when a flood
occurs.  For a variety of depths of floods, and the associated depth
of water in a structure, FIA has data, which it calls the
depth-percent-damage relationship, or damage by elevation (DELV)
values, that provide estimates of the percent of the value of a
structure that is expected to be damaged.  Information is presented
by 1-foot increments of flood level within the structure and
expressed as the average percentage of the property's value that will
be damaged due to a flood of that elevation.  For example, in 1987
DELV information, it was predicted that if water reached a depth of 2
feet within a one-story, no-basement structure located in the AE
zone, 21 percent of the property's value would be damaged, and a
depth of 4 feet of water within the same structure would cause a
29-percent value damage rate. 

In A zones, it is assumed that damage will not begin to occur to a
structure until water reaches the bottom of its lowest floor. 
However, depth damage tables for the VE zone include damage estimates
before water actually reaches the lowest floor of the structure.  In
estimating expected damage to a structure in the V zones, it is
assumed that damage--because of erosion, for example--begins to occur
before water or wave action rises to the level of the structure.\5

As with the PELV data, information used in establishing DELV values
was obtained primarily from engineering studies.  In 1973, data for
DELVs were selected on the basis of studies done by the U.S.  Army
Corp of Engineers and available flood claims at that time. 
Currently, DELV values in the AE zone are updated on the basis of
claims data available from flood insurance policies since 1978.\6


--------------------
\5 These additional risks of damage below the lowest floor of a
structure are included in rate-setting by adding expected damage due
to 20-year events and successively more serious events, each measured
by an additional foot of flood waters, up to the point that water
actually reaches the structure.  Those additional damage estimates
are then factored into the DELVs used in the actuarial rate formula,
which assumes the damage does not begin until water reaches the
lowest floor of the structure. 

\6 FIA determines whether it has sufficient data on floods of
different severities since 1978 to actually replace the original DELV
values.  If data are sufficient, then there is "full credibility,"
and the original DELVs are replaced with DELVs based on experience
under the program since 1978.  If not enough claims data exist for
full credibility, DELVs are based on a weighted average of the
original base table values and the experience data since 1978, where
the weight of the latter is the ratio of actual experience claims to
the number of experience claims necessary for full credibility.  This
would mean that, over time, the original, theoretical DELV values
will have less weight in determining actual DELV values used for
rate-setting, although this will happen much faster in the case of
shallow-depth floods for which data (that is, claims from flood
losses) will accumulate much more rapidly to allow credibility
analysis. 


      EXPECTED DAMAGE ESTIMATES
------------------------------------------------------ Appendix II:3.3

Knowledge of the elevation-frequency relationship and the
depth-damage relationship allows a summing up of the range of flood
probabilities and their associated damage to property and contents.\7
That is, each possible flood is multiplied by the expected damage
should such a flood occur, and each of these multiplications is then
added together.  This summing up of each possible flood's damage
provides an expected per annum percent 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
rating-setting. 

Expressing this mathematically, where i is measured in increments of
1-foot or less, the fundamental concept of rate-setting is: 



Where: 

PELVi is the probability, in a given year, of water surface reaching
or exceeding elevation i, relative to the BFE. 

DELVi is the percentage of property value damage to a structure due
to a flood of elevation i. 

m is the elevation at which flood waters reach the lowest floor of a
structure. 

M is the elevation at which the maximum amount of value damage to the
structure is incurred--floods of a higher elevation are extremely
rare. 

The equation indicates that, for the set of structures of a given
type, in a given zone, and of a given elevation relative to the BFE,
expected damage through flooding is estimated by summing the damage
that could occur to such a structure through a set of possible
flooding events, beginning with a flood that brings waters high
enough to reach the lowest floor of the structure, at elevation m.\8
Increments of 1 foot for successively worse (and less likely) flood
possibilities are then added until the point at which the maximum
probable amount of damage is incurred--elevation M; at that point,
worse floods are extremely unlikely to occur.  Since the damage that
will occur with different types of flooding are multiplied (that is,
weighted) by the probability of a flood of that type occurring, the
summation equals a per annum expected damage (as a percent of value)
due to all possible flooding events. 

Each zone is characterized by different probabilities of floods
occurring (that is, different PELVs), and some have different damage
consequences when a flood of a given elevation does occur (that is,
different DELVs).  Therefore, the formula provides different expected
damage estimates across zones.  Additionally, within a zone, the
estimation is repeated for all different elevations of structures,
relative to the BFE.  Thus, the formula will generate lower rates for
structures elevated above the BFE than for structures elevated to the
BFE because it will take a storm of a greater severity to bring flood
waters to elevation "m" for the more elevated structure.  Finally,
within a zone and for a given elevation of structure relative to the
BFE, the calculation is repeated for several categories of
structures.  For example, structures with basements generally pay
more than those without them, and structures with more than one floor
above basement level generally pay less than those with only one
floor above the basement level. 

The formula shown above will provide a rate per $100 of purchased
insurance.  The rate is then multiplied by how many hundreds of
dollars of insurance coverage are being purchased to determine the
premium for an individual policy.  Several other considerations about
the per $100 rate, as well as policy fees, need to be considered,
however.  The next section discusses these additional issues. 


--------------------
\7 The method of summation approximates calculating the area under a
curve through integration.  The estimation approximates the area
defined by a function that expresses expected damage due to floods
that occur with different probabilities. 

\8 In the AE zone, the elevation of a structure is measured at the
top of the finished flooring of the lowest floor, while in the VE
zone, it is measured at the bottom of the floor beam below the lowest
floor. 


   OTHER COMPONENTS OF ACTUARIAL
   RATES
-------------------------------------------------------- Appendix II:4

The formula for actuarial rates discussed in the previous section is
not the complete actuarial rate-setting formula.  Mathematically, the
more complete formula is: 



Here, additional variables are included to take into consideration
several issues or effects that are important for modifying expected
losses or for building additional expense items into the rates.  The
rest of this section describes each of these additional variables. 


      THE LOSS ADJUSTMENT
      FACTOR--LADJ
------------------------------------------------------ Appendix II:4.1

Rates are "loaded," or adjusted upwards, by approximately 4.2 percent
to account for costs associated with claims and loss adjustment. 
This is called the loss adjustment factor, or LADJ, in the actuarial
formula.  Data on previous years' costs for these tasks are used to
develop the LADJ factor, which can change from the 1993 level of 4.2
over time. 


      THE DEDUCTIBLE OFFSET--DED
------------------------------------------------------ Appendix II:4.2

Currently, the deductible is $500 for most actuarial policies.\9 This
means that the first $500 of any claim that is filed is not covered
under the program's policies.  The fact that some portion of each
claim will not be covered needs to be taken into account so that
rates can be adjusted downward to reflect a lower risk to the program
in insuring properties for flood loss.  To do this, FIA uses a
formula that converts the dollar level of the deductible, which was
$500 in 1993, into a factor for the rating formula.  This formula is
based on experience data on the degree to which losses have been
reduced due to the deductible, with any adjustments necessary to
account for the current policies in force and inflation.\10
Currently, the deductible factor is approximately .95 for structures
that accommodate one to four families, meaning that rates, per $100
of insured value, are reduced by about 5 percent due to the existence
of the $500 deductible. 


--------------------
\9 Lower flood insurance rates are available if the policyholder
agrees to have a significantly larger deductible.  This type of
policy is not very common, however. 

\10 If the level of the deductible does not change, the rise in the
general price level will cause the percent of damage that the
deductible represents to decline.  That is, a constant deductible
will represent a smaller percentage of expected claims damage over
time.  Therefore, the formula underlying the DED factor takes into
account the effect of inflation. 


      THE UNDERINSURANCE
      FACTOR--UINS
------------------------------------------------------ Appendix II:4.3

The basic (PELV*DELV) relationship implicitly assumes that all
policies are for full insurance, meaning that each policy covers the
full value of the insured property.  However, this may not be the
case.  The fact that people often underinsure causes the risk, per
$100 of insurance premium, to be greater, since claims are more
likely to be made against the first few thousand dollars of insurance
coverage.  Therefore, with underinsurance, the per $100 rate of
insurance needs to be higher than in the full insurance case.  The
UINS factor adjusts rates for the degree to which people, on average,
underinsure.  FIA uses experience data on underinsurance factors and
claims data since 1978 to develop the UINS factor for different zones
and types of structures.  More recent experience is given a greater
weight in determining UINS factors.  According to FIA officials,
rates are currently adjusted upwards by about 20 percent due to
underinsurance. 


      EXPENSE ITEMS--EXLOSS
------------------------------------------------------ Appendix II:4.4

EXLOSS is a factor that loads rates for certain expenses, such as
agents' commissions, certain costs of policy sales, as well as for
contingency costs due, for example, to risk of unknown hazards.  The
factor was equal to .74 in 1992 for the AE zone, so that rates are
increased by over 30 percent due to estimated EXLOSS costs.\11 The
costs accounted for in EXLOSS are those that are related in part to
the amounts of insurance that people are buying and the price (that
is, the rate) for that insurance.\12


--------------------
\11 For V-rated zones, EXLOSS was only .69 in 1992 because FIA builds
in higher contingency costs for these zones. 

\12 Two additional fees are added into premiums.  These fees are not
part of the rate per $100, but rather are added into each policy as
flat fees no matter how much insurance a particular policyholder is
buying.  These fees cover certain expenses spread equally over all
policies, as opposed to risk-related costs.  The "expense constant,"
a $45 per policy charge, recaptures certain costs that are incurred
in writing flood insurance policies.  In addition, the "federal
policy fee," a $25 per policy charge, supports flood insurance
studies, floodplain management activities, and the administration of
the program by the federal government. 


   EXAMPLES OF FLOOD INSURANCE
   RATES
-------------------------------------------------------- Appendix II:5

This section contains examples of flood insurance rates to show how
risk factors considered in the application of the hydrologic model
result in differences in actual flood insurance rates.  As noted
throughout this appendix, rates are set for several categories of
properties defined by the key characteristics of flood risk,
including the zone within which the property lies; the elevation of a
structure relative to the BFE; and the type of structure.  Another
issue, however, has not been previously discussed--rate differences
between rates for "basic" limits coverage and rates for "additional"
limits coverage. 

Basic limits rates apply to the first $45,000 dollars of insurance
that is purchased by a policyholder for a single-family structure. 
If the buyer purchases more than $45,000 of coverage, the additional
limits rates apply on any coverage over that amount.\13 The reason
that rates differ depending on the amount of insurance that is
purchased is that claims are more likely to be made against the first
several thousand dollars of coverage; therefore, rates for basic
limits coverage need to be considerably higher than rates for
additional limits.  The formula discussed earlier, with several
important differences in the treatment of underinsurance
considerations, is used to determine rates on both basic limits and
additional limits.  Since the probability values for floods creating
very high levels of damage are lower, the formula generates rates for
additional limits that are considerably less than basic limits
rates.\14


--------------------
\13 Over time, FIA has adjusted the level of insurance sold at basic
limits rates.  FIA uses experience data to determine the appropriate
level at which rates per $100 of insured value should decline.  In
setting the $45,000 level, FIA examines available data on past claims
and looks for a natural break in the relationship between premiums
and losses.  The rise in the price level over time will cause that
break to rise. 

\14 It would actually be more accurate to have several rate levels
that decline as a higher level of insurance is purchased.  FIA
officials told us, however, that they use only two rate levels for
simplicity, so as not to complicate the work of insurance agents in
pricing insurance for the potential insured. 


      ZONES A1-A30, AE
------------------------------------------------------ Appendix II:5.1

The majority of post-FIRM structures lying within the 100-year
floodplain are in what is currently called the AE zone.  At one time,
there were 30 numbered A zones, each of which was charged different
rates.  Because rate differences across these zones were very slight
for post-FIRM properties, FIA now rates all numbered A zones together
and has renamed the zone AE.  Zone AE has many different post-FIRM
rates, depending on the certain characteristics of the structure.  In
table II.1, we show rates for a one-floor, no-basement, single-family
structures in the AE zone.  In addition, the table shows rates for
building coverage, as opposed to rates for contents coverage (which
are generally higher).  The table shows rates for both basic limits
coverage--coverage up to $45,000--and additional limits, which is the
rate for coverage over $45,000. 



                          Table II.1
           
              AE Zone Premium Rates for Certain
             Structures per $100 of Insured Value

Elevation of lowest                                 Rate for
floor above or below            Rate for          additional
BFE                         basic limits              limits
--------------------  ------------------  ------------------
+3 or more                         $0.14               $0.06
+2                                  0.16                0.06
+1                                  0.21                0.06
0                           0.33                0.06
-1                                  0.86                0.06
------------------------------------------------------------
Note:  These are the rates for a one-floor, single-family building
without a basement. 


      ZONES V1-V30, VE
------------------------------------------------------ Appendix II:5.2

As with the numbered A zones, the original 30 numbered V zones have
been combined and renamed zone VE.  Less than 2 percent of post-FIRM
structures are in the VE zone, which includes coastal areas subject
to wind and wave action.  In table II.2, we show rates for a
structure in the V zone that is in compliance with the program's
current building standards for the V zone.  In the V zone, rather
than having basic limits and additional limits rates, there are three
sets of rates; the rates depend on the degree of underinsurance,
relative to the replacement cost of the structure, that the
policyholder has purchased.  FIA uses this rate structure for the V
zone because rates are high, and people have a greater tendency to
underinsure.  The rate structure in V zones gives people an incentive
to insure more fully, since rates per $100 of insured value are lower
the greater the coverage one buys.  The table shows that rates in the
V zone are substantially higher than in the A zone. 



                          Table II.2
           
              VE Zone Premium Rates for Certain
             Structures per $100 of Insured Value


Elevation of   75 percent or        50 to 74        Under 50
lowest floor         more of      percent of      percent of
above or         replacement     replacement     replacement
below BFE               cost            cost            cost
------------  --------------  --------------  --------------
+4 or more             $0.36           $0.48           $0.71
+3                      0.41            0.56            0.83
+2                      0.53            0.71            1.06
+1                      0.71            0.96            1.35
0               0.93           $1.25            1.69
-1                     $1.23           $1.63           $2.12
------------------------------------------------------------
Note:  These rates are for a one-floor, single-family building
without a basement. 


      OUTSIDE THE IDENTIFIED
      100-YEAR FLOODPLAIN
------------------------------------------------------ Appendix II:5.3

Most properties not in the identified 100-year floodplain are in zone
X, which includes properties outside of a special flood hazard area,
or the 100-year floodplain.\15 One of the most important differences
in rates in the X zone is that they are not set using the hydrologic
model.  Rates are the same across the entire zone for a given type of
building, since BFEs are not defined within the zone.  Table II.3
gives examples of building rates in the X zone for single-family
structures with no basement. 



                          Table II.3
           
               Zone X Premium Rates for Certain
             Structures per $100 of Insured Value

Basic limits                   Additional limits
-----------------------------  -----------------------------
$0.25                          $0.07
------------------------------------------------------------
Note:  These rates are for a one-floor, single-family structure
without a basement. 


--------------------
\15 Zone X includes areas that used to be classified as either zone B
or C. 


MAJOR CONTRIBUTORS TO THIS REPORT
========================================================= Appendix III

RESOURCES, COMMUNITY, AND ECONOMIC
DEVELOPMENT DIVISION, WASHINGTON,
D.C. 

Robert S.  Procaccini, Assistant Director
Eugene J.  Chuday, Jr., Evaluator-in-Charge
Amy D.  Abramowitz, Senior Economist

ACCOUNTING AND INFORMATION
MANAGEMENT DIVISION, WASHINGTON,
D.C. 

Terry L.  Carnahan, Assistant Director
