Consumer Impacts of Property Insurance Surcharges in Florida

Consumer Impacts of Property Insurance Surcharges in Florida

Prepared by Staff of the Florida House of Representatives Committee on Financial Services
October 24, 1997

Table of Contents

Part I: Background: Residential Property Insurance in Florida
How does homeowners' insurance work?
Statutory principles of property insurance ratemaking
The Florida property insurance market

Part II: State-created Insurance Entities
Residual property insurance markets: Residential Property and Casualty Joint Underwriting Association (RPCJUA)
Residual property insurance markets: Florida Windstorm Underwriting Association (FWUA)
Florida Hurricane Catastrophe Fund (FHCF)

Part III: Subsidies in Residual Market Assessments
The recoupment process
Residual market assessments as subsidies
Future assessment scenarios
Loss scenario 1: Expected annual loss
Loss scenario 2: 5-year probable maximum loss
Loss scenario 3: 10-year probable maximum loss
Loss scenario 4: 20-year probable maximum loss
Loss scenario 5: 50-year probable maximum loss

Part IV: Subsidies and Catastrophe Fund "Recoupment"
Statutory treatment of Catastrophe Fund premiums
The impact of uniform percentage surcharges
Catastrophe Fund "recoupment" surcharges in the marketplace
Catastrophe Fund emergency assessments

Part V: Implications of Subsidized Insurance Costs
The arguments for subsidies
The arguments against subsidies

Conclusion

List of Tables

Table 1: Country-wide premiums and combined ratios for selected lines of property and casualty insurance, 1990-1995

Table 2: Total value of residential properties insured under policies providing windstorm coverage as of September 30, 1996

Table 3: Aggregate statewide property insurance premiums in Florida, 1993-1995

Table 4: Florida's population, 1990-2000

Table 5: Share of insured value covered by the residual market, 6 largest counties, personal lines and commercial lines residential, as of September 30, 1996

Table 6: RPCJUA personal lines residential policy count and exposure as of September 30, 1997, excluding policies that do not provide windstorm coverage

Table 7: Expected annual RPCJUA hurricane losses by county

Table 8: FWUA policies in force and exposure as of August 31, 1997

Table 9: Comparison of key features of the RPCJUA and the FWUA

Table 10: The role of the Florida Hurricane Catastrophe Fund in a hypothetical hurricane comparable to Hurricane Andrew

Table 11: Examples of premiums paid by insurers to the Florida Hurricane Catastrophe Fund for each $100,000 of personal lines residential exposure with a $500 deductible

Table 12: Premium examples for a homeowners' policy covering a $75,000 frame house, 15 largest counties

Table 13: Hurricane loss projections used in residual market assessment scenarios

Table 14: Summary of deficit assessments in 5 hurricane loss scenarios

Table 15: Summary of hypothetical example of uniform percentage surcharges for Catastrophe Fund premium recoupment

Consumer Impacts of Property Insurance Surcharges in Florida

Executive Summary

The purpose of this report is to describe the circumstances in which residential property insurance consumers in Florida may be required to pay surcharges which support the insurance costs of other Florida consumers. The report is intended to provide legislators with information on these questions:

To what extent are subsidies relied on to fund the residential property insurance system in Florida?
How are the subsidy costs distributed among Florida consumers?

Part I of the report summarizes basic principles and facts about residential property insurance and the Florida residential property insurance market. The key principles are:

Property insurance is a risk-transfer mechanism, under which an insurance company assumes specified risks that would otherwise be borne by the homeowner.
The amount paid by a homeowner for property insurance should reflect the likelihood that the homeowner will generate a claim during the term of the insurance policy.
By law, property insurance rates are "inadequate" if they are clearly insufficient to sustain projected losses and expenses.
By law, property insurance rates are "unfairly discriminatory" if surcharges are applied in a manner that does not reflect differences in various policyholders' expected losses.

In the years since Hurricane Andrew, Florida has suffered a crisis in the availability of residential property insurance. Statewide, almost 15 percent of residential properties are insured by a state-created insurer of last resort. More than 30 percent of Dade County residential properties, and more that 25 percent of Broward County residential properties are covered by insurers of last resort.

Florida's growth contributes to problems of availability. Florida's population is growing by approximately 250,000 persons a year, and the number of housing units in Florida is growing by a net amount of approximately 150,000 new units a year. In addition to providing coverage for existing homes, the capacity of the property insurance system must grow in order to cover these new homes.

Part II of the report provides information on the structure and funding of three state-created property insurance entities: the Residential Property and Casualty Joint Underwriting Association (RPCJUA), the Florida Windstorm Underwriting Association (FWUA), and the Florida Hurricane Catastrophe Fund (FHCF).

The RPCJUA and the FWUA are state-created insurers of last resort, composing what is referred to as the "residual" property insurance market. The Legislature has required that the RPCJUA and the FWUA collect adequate premiums from their policyholders. When the premiums and other revenues of the RPCJUA or the FWUA are not sufficient to pay their claims, they have the power to cover their deficits by levying assessments on insurance companies and their policyholders. The RPCJUA has levied a $40.5 million assessment to cover its 1995 deficit, and the FWUA has levied a $117 million assessment to cover its 1995 deficit.

The Florida Hurricane Catastrophe Fund is a state-created reinsurance pool administered by the State Board of Administration. The fund reimburses insurance companies for a portion of their claims in a major hurricane, and collects approximately $450 million a year from insurance companies in exchange for the promised reimbursement. The fund's premium structure is designed to reflect the relative likelihood of various geographic areas to sustain economic loss from major hurricanes. When the balance in the fund is insufficient to meet the fund's obligations, the fund has the power to issue bonds and levy assessments of up to 4 percent on all property and casualty insurance premiums other than workers' compensation premiums (that is, not only property insurance, but also automobile, commercial multi-peril and liability, medical malpractice, professional liability, and other forms of insurance classified as "property and casualty"). The fund currently has a balance of approximately $2 billion and the ability to bond for an additional $5.5 billion.

Part III of the report addresses the subsidies contained in RPCJUA and FWUA assessments. As authorized by law, an insurance company collects the assessments from its consumers as a uniform percentage of the consumer's total property insurance premium. The result is that the assessment burden, which related directly to hurricane losses, is not distributed among consumers in a way that reflects various consumers' relative vulnerability to hurricane losses.

Part III presents projections of residual market assessments in several hurricane loss scenarios: the expected annual loss, and the 5-year, 10-year, 20-year, and 50-year probable maximum losses. Each of these scenarios would lead to assessments by both the RPCJUA and the FWUA, except that the RPCJUA would not need to levy an assessment in the case of its 5-year probable maximum loss. Losses larger than the 10-year probable maximum loss would require assessments stretching over more than one year.

Part IV addresses the means by which premiums paid by insurers to the Catastrophe Fund are reflected in amounts the insurers charge their policyholders. Each insurer's Catastrophe Fund premium is based on an actuarial analysis of that insurer's portfolio of exposures. The residual market and several large insurance companies are using a uniform percentage surcharge on all of their residential property insurance premiums to "recoup" their Catastrophe Fund costs from their policyholders. The percentage surcharge method results in some consumers paying more to their insurance company, and some consumers paying less, than the Catastrophe Fund charged the insurance company for that particular consumer's exposure.

The bonding authority of the Catastrophe Fund also creates subsidies. The Catastrophe Fund's emergency assessment powers allow for a situation in which many lines of insurance subsidize property insurance costs.

Part V outlines the major arguments for and against subsidy of property insurance costs. Arguments in favor of subsidies include:

Subsidies are needed as a substitute for surplus in the residual market.
Subsidies are needed to assure "affordability" of residual market premiums.
Regardless of residual market premium levels, there will be some disasters, of the magnitude of Hurricane Andrew or greater, in which subsidies are the only alternative to large numbers of unpaid claims.
The assessment powers of the Catastrophe Fund would be exercised only after a disaster so large as to threaten many property and casualty insurers' solvency. Assessment of many lines of insurance is necessary to protect the property and casualty insurance industry's ability to write many lines of insurance.

Arguments against subsidies include:

Subsidies unfairly burden low-risk homeowners for the benefit of high-risk homeowners.
Subsidies encourage high-risk coastal development by masking the market costs of coastal development.
Subsidies remove an incentive for loss mitigation.
Subsidies encourage growth of the residual market, thereby encouraging further subsidies.

Introduction

The crisis in availability and affordability of residential property insurance in Florida is a direct result of the most costly natural disaster in American history, Hurricane Andrew. Since Andrew struck southern Dade County in 1992, causing more than $16 billion of insured losses, property insurance premiums have risen dramatically and property insurance has, in many parts of the state, become difficult to obtain through ordinary means.

Florida has two insurers of last resort, which were intended to assure the continued availability of property insurance coverage when property owners could not obtain the coverage through traditional means. Today, these two insurers of last resort provide windstorm coverage for over 900,000 Florida properties.

Florida also has a state-operated reinsurance pool, designed to preserve the ability of insurance companies to pay claims after major disasters. Today, the reinsurance pool has a fund balance of approximately $2 billion, and the ability to bond for another $5.5 billion.

All of these entities are funded to some extent by payments from property owners. In some situations, some Floridians' insurance premiums are surcharged to cover the costs of insuring other Floridians.

To what extent are subsidies relied on to fund the residential property insurance system in Florida?
How are the subsidy costs distributed among Florida consumers?

Part I: Background: Residential Property Insurance in Florida

How does homeowners' insurance work?

A homeowners' insurance policy is a contract under which an insurance company promises that it will pay certain amounts if certain events occur within a specified period of time. As with other forms of property and casualty insurance, such as motor vehicle insurance or commercial liability insurance, homeowners' insurance is not a "risk-sharing" mechanism in which all homeowners are responsible for each homeowner's losses, but is instead a risk-transfer mechanism: for a specified price and with specified limitations, an insurance company assumes the risk that would otherwise be borne by the homeowner.

The statutory definition of "insurance," s. 624.02, F.S., provides: "'Insurance' is a contract under which one [i.e., the insurer] undertakes to indemnify [i.e., provide compensation for damages incurred to] another [i.e., the insured] or pay or allow a specified amount or a determinable benefit upon determinable contingencies." A contract under which one party agrees to pay losses sustained by another party is, as to those specified losses, an assumption by one party of the other party's risk.

The basic principle of property and casualty insurance is that if the insurer enters into a large enough number of these contracts and charges enough for them, current revenues will be enough to enable the insurer to cover its expenses and pay all of its claims.

Revenues are also needed to secure access to capital resources in order to pay claims from the foreseeable, but extreme, event, such as Hurricane Andrew. Insurance companies were able to pay $15.5 billion of the $16 billion in insured losses from Hurricane Andrew, even though the claims payments equaled about 10 years' worth of Florida property insurance premiums (not just the premium attributable to hurricanes, but the entire premium collected for fire, theft, liability, windstorm, and other perils). The remainder of the claims, which were against policies issued by companies that became insolvent, were paid by the Florida Insurance Guaranty Association.

Companies were able to pay the claims because a portion of their premium revenues had been used to secure access to capital. The most common way companies gain access to the funding needed to pay catastrophic claims is by obtaining reinsurance. Reinsurance provides payments to insurance companies when they sustain specified levels of losses; many companies used premium revenues to procure the reinsurance that funded their Andrew claims and continue to procure the reinsurance needed to pay claims in the next major hurricane. Many companies sustained substantial reductions in their surplus (net worth) in order to pay the claims and continue to commit substantial amounts of surplus to pay claims in the next major hurricane. Other mechanisms, such as lines of credit, revenue bonds, and catastrophe futures have evolved to provide access to capital to pay catastrophic claims, but all of these risk-transfer and risk-deferment mechanisms entail costs that must be covered out of an insurer's revenues.

In essence, a homeowners' insurance policy is a wager. The insurance company bets that an event (a fire, for example) will not occur, and the homeowner bets that it will occur. If the event occurs, the insurance company must pay a large amount of money to the "winning" bettor; the bettor must pay a lesser amount up front ("premium") in order to make the wager.

Wagering systems find ways to account for the relative chance that a particular event will or will not occur. A $2 bet on a favorite in a horse race will not, if the horse wins, pay as much as a winning $2 bet on a longshot. A football bet is a winner not because the bettor's team won the game, but because the bettor's team beat a point spread. As with the odds in a horse race, the spread on a sports bet is set by oddsmakers to reflect the relative probability that a particular bet will be a winner.

A Las Vegas casino's sports book assumes the risk of paying off all winning bets. The revenues generated by all of the losing bets provides the funding system to pay off all of the winning bets, but the sports book would have a duty to pay the winners even if the losing bets did not generate enough revenue. When the point spread is properly set, the system is likely to work; that is, properly set odds should enable the casino to pay off bets without losing money itself. When the oddsmaker errs, it is more likely that the casino will lose money.

The casino is willing to assume the risk of paying off winning bets if it is confident that each individual bet will accurately reflect the relative likelihood that it will be a winner. Oddsmakers try to use the most accurate and relevant information to calculate the relative likelihood that a particular bet will be a winner, and use the point spread to price the bet in accordance with that relative likelihood.

Property and casualty insurance operates on similar principles. The insurer is willing to be on the receiving end of a risk-transfer if it is confident that each premium payment from a homeowner will accurately reflect the likelihood that the insurer will have to pay a particular claim to that particular homeowner within the policy period. Not unlike Las Vegas oddsmakers, insurance actuaries try to use the most accurate and relevant information to calculate the relative likelihood that a specific insured risk will generate a claim in a given time.

Everyone who has purchased automobile insurance is familiar with the process used to assure that premiums are matched to the likelihood of loss. To determine the premium that the insured will pay, the insurer uses information about the vehicle, where and how far the vehicle is driven, the driving records of the people who will drive the vehicle, and other factors (such as the presence of teenage drivers in the household) that affect the relative likelihood of claims. Similarly, homeowners' insurance premiums have always reflected the relative likelihood of fires or thefts. Masonry homes generally have lower premiums than wood frame homes of similar value (assuming that the other risk factors are equal for both homes) because masonry homes are more fire-resistant. All else being equal, a home that is protected by a city fire department will usually have lower premiums than a home that relies on a volunteer fire department, and a home in a low-theft area will usually have lower premiums than a home in a high-theft area.

Controversies have arisen over the application of these principles to projected catastrophic windstorm losses. Unlike fire or theft losses, catastrophic windstorm losses vary greatly from one year to the next, and, while it is possible to arrive at reasonable projections of an average year's windstorm losses or of the largest windstorm loss that an insurer can reasonably expect, there is no known way to project with any credibility the catastrophic windstorm losses for a particular specified year. Methods of projecting average losses and maximum losses over time have also been controversial. The Legislature created the Florida Commission on Hurricane Loss Projection Methodology under section 627.0628, F.S., to provide an independent, expert review of computer models used to project hurricane losses. The findings of the commission have not been universally accepted.

The analogy between insurance companies and gambling operations goes only so far. While gambling operations adjust the odds to assure that a small portion of each bet goes to the "house" as profit, insurance companies rely on investments to generate profits. In a typical year, the claims and expenses paid by an insurance company will exceed the revenues received from consumers.

Table 1 shows the insurance industry's country-wide premiums and profitability for four lines of insurance (private passenger auto liability, private passenger auto comprehensive and collision, commercial multi-peril, and homeowners' multi-peril) for the years 1990 through 1995. "Combined ratio" is the ratio of an insurance company's claims and expenses to its premiums. A combined ratio of 105, for example, means that an insurance company paid out $105 in claims and expenses for every $100 it received in premiums.

Table 1: Country-wide premiums (in millions ) and combined ratios for selected lines of property and casualty insurance, 1990-1995*

1990 1991 1992 1993 1994 1995

Private passenger auto liability

Premiums written $47,831 $51,200 $55,473 $59,273 $61,952 $65,337

Combined ratio** 117.7 113.3 109.3 107.7 104.9 101.8

Private passenger auto comprehensive and collision

Premiums written $30,562 $31,555 $32,898 $34,102 $34,861 $36,639

Combined ratio** 93.9 88.5 87.7 88.9 92.6 97.2

Commercial multi-peril

Premiums written $17,709 $17,032 $16,432 $17,308 $17,801 $18,845

Combined ratio** 107.6 110.8 126.7 115.5 118.7 112.4

Homeowners' multi-peril

Premiums written $18,577 $19,303 $20,477 $21,546 $21,551 $23,987

Combined ratio** 112.5 117.3 158.0 113.1 118.0 112.3

* Source: Data from A. M. Best Co., as published in The Fact Book 1997: Property/ Casualty Insurance Facts, Insurance Information Institute, 1996.
** "Combined ratio" is the ratio of losses (claims) and expenses to premiums.

Statutory principles of property insurance ratemaking

Amounts charged for property insurance are regulated by the Department of Insurance under sections 627.062 and 627.0629, Florida Statutes. The general requirement is set out in subsection (1) of section 627.062, F.S.:

627.062 Rate standards.--

(1) The rates for all classes of insurance to which the provisions of this part are applicable shall not be excessive, inadequate, or unfairly discriminatory.

The statute goes on to define the terms "excessive," "inadequate," and "unfairly discriminatory."

Property insurance rates are excessive if:

They are likely to produce a profit from Florida business that is unreasonably high in relation to the risk involved (s. 627.062(2)(d)1., F.S.);
Expenses are unreasonably high in relation to services rendered (s. 627.062(2)(d)1., F.S.); or
The rate structure provides for replenishment of surplus (net worth) from premiums, where the replenishment is attributable to investment losses (s. 627.062(2)(d)2., F.S.). This means that a company cannot raise premiums to make up for a bad year on Wall Street.

Property insurance rates are inadequate if:

They are clearly insufficient, together with investment profits attributable to them, to sustain projected losses and expenses in the class of business to which they apply (s. 627.062(2)(d)3., F.S.); or
Discounts or credits are allowed which exceed a reasonable reflection of expense savings and reasonably expected loss experience from the risk or group of risks receiving the discount or credit (s. 627.062(2)(d)5., F.S.).

Property insurance rates are unfairly discriminatory if:

The application of premium discounts, credits, or surcharges among various risks does not bear a reasonable relationship to the expected loss and expense experience of the risks (s. 627.062(2)(d)6., F.S.).

Section 627.0629 sets out several forms of discounts that are allowed or required in property insurance rate plans.

A rate filing must include appropriate discounts, credits, or other rate differentials, or appropriate reductions in deductibles, for properties on which fixtures that reduce the amount of loss in a windstorm (for example, hurricane shutters) have been installed (s. 627.0629(1), F.S.).
A rate filing must include positive and negative rate factors that reflect the manner in which building code enforcement in a particular jurisdiction addresses risk of wind damage (s. 627.0629(2), F.S.).
A rate filing for mobile home owners' insurance must include appropriate discounts, credits, or other rate differentials for mobile homes that meet the 1994 U.S. Department of HUD standards and that also meet applicable tie-down requirements (s. 627.0629(3), F.S.).
An insurer may provide discounts of up to 10% for a mobile home if the policyholder provides evidence of a current inspection and certification that tie-downs have been properly installed and are in good condition (s. 627.0629(8), F.S.).

The Florida property insurance market

As of September 30, 1996, $789 billion worth of residential property was insured by some form of property insurance policy that included windstorm coverage. Table 2 shows the total residential insured values for the six largest counties and the state as a whole. Coverage is broken down by the type of policy: "personal lines residential" refers to all coverage of site-built one-to-four family homes; "commercial lines residential" refers to condominium associations, apartment buildings, and similar risks; and "mobile home" refers to mobile homes and other manufactured housing.

Table 2: Total value of residential properties insured under policies providing windstorm coverage as of September 30, 1996 (in millions)*

County Personal lines residential Commercial lines residential Mobile home Total insured value County percentage of statewide total insured value

Dade $74,339 $9,037 $253 $83,629 10.6%

Broward $71,980 $10,101 $733 $82,815 10.5%

Palm Beach $71,771 $8,923 $557 $81,251 10.3%

Pinellas $43,020 $5,534 $1,655 $50,209 6.4%

Orange $39,597 $3,351 $693 $43,642 5.5%

Hillsborough $39,373 $2,494 $1,095 $42,963 5.4%

All others $349,697 $32,082 $22,610 $404,386 51.2%

State total $689,777 $71,522 $27,596 $788,895

* Source: Florida Hurricane Catastrophe Fund

Floridians pay over $3 billion a year in residential property insurance premiums. Table 3 shows the statewide aggregate property insurance premiums for 1993, 1994, and 1995. Data are broken down into "personal lines," which includes both personal lines residential as described above and mobile homes, and "commercial lines residential" as described above. The table also distinguishes between premiums paid to insurance companies ("voluntary market") and premiums paid to the state-created insurers of last resort, the Residential Property and Casualty Joint Underwriting Association and the Florida Windstorm Underwriting Association.

Table 3: Aggregate statewide property insurance premiums in Florida, 1993-1995 (in millions)*

Year Personal lines residential Commercial lines residential

Voluntary market Residual market Total Voluntary market Residual market Total

1993 $1,466 $116 $1,582 $722 $17 $739

1994 $1,670 $376 $1,976 $796 $51 $847

1995 $1,824 $449 $2,273 $890 $51 $941

*As calculated by the Residential Property and Casualty Joint Underwriting Association (RPCJUA) for purposes of determining their assessment base, published in the Offering Memorandum for the RPCJUA's debt issue, May 5, 1997.

The growth in premiums over the years is attributable in part to increased premium rates and in part to Florida's continuing population growth. According to the Economic and Demographic Research Division of the Joint Legislative Management Committee, Florida's population is growing at the rate of approximately 250,000 persons a year. More significantly, the Economic and Demographic Research Division has estimated that Florida's housing stock grows by approximately 150,000 net new units a year, all of which will need to be insured under property insurance policies providing windstorm coverage.

Table 4 shows Florida's population growth. Population is shown for the state as a whole and for the six largest counties in 1990, 1995 (estimated), and 2000 (projected). The percentage changes from 1990 through 2000 are also shown.

Table 4: Florida's population, 1990-2000 (in thousands)*

County 1990 1995 2000 Percentage change, 1990-2000

Dade 1,937 2,005 2,118 9.9%

Broward 1,255 1,355 1,453 15.8%

Palm Beach 863 956 1,059 22.7%

Hillsborough 834 887 952 14.1%

Pinellas 851 874 906 6.5%

Orange 677 752 830 22.6%

All others 6,521 7,224 8,005 22.8%

State total 12,938 14,053 15,323 18.4%

*Source: Division of Economic and Demographic Research, Joint Legislative Management Committee

A substantial number of Florida homes are covered by state-created insurers of last resort, which compose the "residual market." The insurance companies that are licensed (or "admitted") to do business in Florida compose the remainder of the market, known as the "voluntary market." Table 5 shows the value of properties insured by the residual market and by the voluntary market for the state's six largest counties and for the state as a whole. The table is based on exposures as of September 30, 1996, the last date for which complete county-by-county exposure information is available. As is discussed later in this report, residual market exposures have changed considerably since that date.

Table 5: Share of insured value covered by the residual market, 6 largest counties, personal lines and commercial lines residential, as of September 30, 1996*

County RPCJUA FWUA Voluntary market

Exposure (in millions) Percentage of county total Exposure (in millions) Percentage of county total Exposure
(in millions)
Percentage of county total

Dade $13,953 16.7% $13,340 15.9% $56,336 67.4%

Broward $12,256 14.8% $8,738 10.6% $61,821 74.6%

Palm Beach $11,958 14.7% 0 0 $69,293 85.3%

Pinellas $4,822 9.6% $2,435 4.8% $42,952 85.6%

Orange $1,560 3.6% 0 0 $42,082 96.4%

Hillsborough $3,561 8.3% 0 0 $39,402 91.7%

State total $66,351 8.4% $45,928 5.8% $676,616 85.8%

* Overall exposure data provided by Florida Hurricane Catastrophe Fund; residual market exposure data provided by Residential Property and Casualty Joint Underwriting Association (RPCJUA) and Florida Windstorm Underwriting Association

Availability problems are greatest in the areas in which the largest proportion of risks are covered by the residual market. Table 5 indicates that availability problems are most severe in Dade and Broward Counties. The growth of the FWUA in Palm Beach County in 1997 (see Table 8 on page 26) indicates that availability problems are also severe in that county. Table 5 also indicates that availability problems are far less severe in inland areas, such as Orange County, and in lower-risk coastal areas, such as Hillsborough County.

Part II: State-created Insurance Entities

Residual property insurance markets:
Residential Property and Casualty Joint Underwriting Association (RPCJUA)

After Hurricane Andrew many insurance companies attempted to reduce their presence in the Florida residential property insurance market or chose not to expand at a rate sufficient to accommodate Florida's growth. These actions created an availability crisis in residential property insurance. One of the Legislature's responses to the availability crisis was the creation of an insurer of last resort, the Residential Property and Casualty Joint Underwriting Association (RPCJUA) under section 627.351(6), Florida Statutes.

The RPCJUA provides personal lines and commercial lines residential property insurance policies to applicants who are unable to obtain coverage from an insurance company. The RPCJUA charges premiums for the coverage it provides, but when the premiums and other resources of the RPCJUA are not sufficient to pay claims, the RPCJUA has the power to levy assessments on insurance companies and their policyholders to generate the revenues necessary to cover the deficit. The RPCJUA has the power to issue bonds and other debt instruments, and to pledge its premiums, assessments, and other resources to pay off the debt.

The RPCJUA statute sets out the requirements for RPCJUA rates. Paragraph 627.351(6)(d), F.S., provides, in part, that:

It is the intent of the Legislature that the rates for coverage provided by the association [i.e., the RPCJUA] be actuarially sound and not competitive with approved rates charged in the admitted voluntary market, so that the association functions as a residual market mechanism to provide insurance only when the insurance cannot be procured in the voluntary market....Rates shall include an appropriate catastrophe loading factor that reflects the actual catastrophic exposure of the association and recognizes that the association has little or no capital or surplus....

This statutory provision establishes several standards for RPCJUA ratemaking:

Rates must be actuarially sound. This means that the rates may not be inadequate, as determined according to the principles of actuarial science.
Rates may not be competitive with the voluntary market. This means that RPCJUA rates may not be lower than rates for comparable coverage that have been approved by the Department of Insurance and are charged by insurance companies. One of the purposes of this requirement is to assure that an applicant does not attempt to obtain RPCJUA coverage because it is less expensive than coverage otherwise available to the applicant from an insurance company.
Rates must include an appropriate catastrophe factor that reflects the RPCJUA's actual catastrophic exposure and recognizes the RPCJUA's lack of any substantial surplus. This means that RPCJUA ratemaking must take into account the RPCJUA's actual concentration of risks. As with other residual market mechanisms, such as the Florida [Automobile] Joint Underwriting Association, it is to be expected that the RPCJUA will cover a higher proportion of "bad" risks--that is, highly vulnerable properties--than a typical insurance company. This requirement mandates that the RPCJUA's rates reflect the RPCJUA's greater likelihood to sustain catastrophic losses.

Paragraph 627.351(6)(d), F.S., also requires that in any given county the average RPCJUA rate be no lower than the highest average rate charged in that county by one of the state's top 20 residential insurers (no lower than the highest of the top 8 in the case of mobile home rates). The purpose of this requirement is to provide an objective test to assure that the RPCJUA is in compliance with the other applicable statutory rate standards.

These rate standards may also be viewed as the Legislature's attempt to balance the RPCJUA's funding sources between RPCJUA policyholders and voluntary market policyholders. The standards may be construed to reflect a policy decision that RPCJUA policyholders should be responsible, through premium payments, for the expected losses of the RPCJUA, but that in extreme cases the public should also provide a portion of the RPCJUA's funding. An assessment system is the means by which non-RPCJUA policyholders provide funding for the RPCJUA.

Subparagraph 627.351(6)(b)3.,F.S., describes the assessment procedures that are used when the RPCJUA sustains a deficit. The RPCJUA is divided into two accounts, one for personal lines residential coverage and one for commercial lines residential coverage. When a deficit occurs in one of the accounts, assessments are levied on insurers writing the type of insurance covered by the account in proportion to their market share (that is, if a deficit occurs in the personal lines account, insurers are assessed in proportion to their personal lines property insurance premiums).

There are two kinds of assessments: "regular" assessments and "emergency" assessments. Regular assessments are levied on insurance companies, which are then allowed to pass the assessment amounts on to their policyholders. Some insurers have been granted exemptions from regular assessments as incentives for them to take over RPCJUA policies. Emergency assessments are levied on policyholders; insurance companies collect emergency assessments and transfer the collections to the RPCJUA. Exemptions from emergency assessments are prohibited.

Subparagraph 627.351(6)(b)3., F.S., imposes several limitations on the amount of assessments.

When the deficit for a particular year is less than 10% of the aggregate statewide voluntary market premium for the subject lines of business (i.e., the types of insurance covered by the RPCJUA account which sustained the deficit), the entire amount of the deficit must be recovered through a one-time regular assessment.
For larger deficits, the RPCJUA must levy both regular and emergency assessments. In that case, the regular assessment would be a one-time assessment that cannot exceed the greater of 10% of the deficit or 10% of the aggregate statewide voluntary market premium for the subject lines of business.
In the case of the larger deficits, the regular assessments would be accompanied by emergency assessments. Emergency assessments are a uniform percentage of the property insurance premium, to be collected by the insurance company upon issuance or renewal of a policy. The aggregate amount of the emergency assessments levied in a particular year may not exceed the greater of 10% of the original deficit plus financing costs or 10% of the aggregate statewide premium for the subject lines of business for the prior year. The emergency assessments continue until the debt incurred by the RPCJUA is retired.

In its early years, the RPCJUA grew rapidly, eventually reaching a peak of more than 935,000 policies in force. The Legislature authorized the RPCJUA to provide incentives to encourage insurers to take over RPCJUA policies (see ss. 627.351(6)(g)3. and 627.3511, F.S.). As a result of these incentives, more than 476,000 RPCJUA policies have been taken over by the voluntary market, and efforts to place additional policies in the voluntary market continue.

In 1996, the RPCJUA levied regular assessments of $40.5 million to cover its accumulated deficit through year-end 1995. The 1996 assessment represented approximately 2.2% of the RPCJUA's assessment base.

The RPCJUA ended calendar year 1996 in the black, with net income of $56,816,470 (calculated on a Generally Accepted Accounting Principles basis); this net amount included $40.5 million in proceeds from the assessments levied in 1996.

Table 6 shows the current status of the RPCJUA. "Structure exposure" is the total value of structures insured by the RPCJUA. "Contents exposure" is the amount of contents coverage (for such non-structural items as furniture and appliances) provided under RPCJUA policies. Data are shown for the five counties that have the highest number of RPCJUA policies and for the state as a whole.

Table 6: RPCJUA personal lines residential policy count and exposure as of September 30, 1997, excluding policies that do not provide windstorm coverage*

Policies in force Premium written (in millions) Structure exposure (in millions) Contents exposure (in millions)

Dade County 142,575 $152 $13,316 $5,571

Dade County as a percentage of state total 25.0% 32.5% 28.5% 27.8%

Broward County 98,320 $99 $9,681 $4,587

Broward County as a percentage of state total 17.3% 21.1% 20.7% 22.9%

Palm Beach County 64,335 $67 $7,354 $3,572

Palm Beach County as a percentage of state total 11.3% 14.3% 15.7% 17.8%

Pinellas County 50,473 $29 $3,362 $1,410

Pinellas County as a percentage of state total 8.9% 6.2% 7.2% 7.0%

Hillsborough County 30,376 $17 $2,102 $820

Hillsborough County as a percentage of state total 5.3% 3.6% 4.5% 4.1%

Top 5 counties 386,079 $364 $35,815 $15,959

Top 5 counties as a percentage of state total 67.8% 77.8% 76.7% 79.6%

State total 569,684 $468 $46,704 $20,039

Source: RPCJUA

Table 6 indicates that the business of the RPCJUA is concentrated in the state's largest counties. The five largest counties represent 44.4% of the state's population (see Table 4 on page 15), but they represent 76.7% of the property values insured by the RPCJUA and 67.8% of the number of RPCJUA policies in force. Further analysis indicates that the RPCJUA is especially concentrated in Dade, Broward, and Palm Beach Counties. Although these three counties account for 31.3% of the state's population and 31.4% of the statewide total value of insured residential properties (see Table 2 on page 13), they account for 65.0% of the RPCJUA's total insured value and 53.6% of the RPCJUA's policies in force.

The RPCJUA relies on a modeling firm, Eqecat Catastrophe Management, for projections of its expected annual hurricane losses and its probable maximum hurricane losses. "Expected annual loss" is the average of the hurricane loss for all hurricane years (based on 10,000 simulated years of hurricanes). "Probable maximum loss" is the highest loss that the RPCJUA is projected to incur at a given probability; for the 100-year probable maximum loss, the probability that the loss would be exceeded in any given year is 1%. Insurers are generally considered solvent if they have resources sufficient to sustain their 100-year probable maximum loss, although a leader in the evaluation of insurance companies, the A. M. Best Co., has suggested that a 250-year standard is more appropriate for insurers with substantial Florida exposures. In general, the claims-paying ability rating of an insurer that has resources to cover the 250-year probable maximum loss will be more favorable than the ratings of an insurer that has resources only for the 100-year probable maximum loss.

According to an analysis prepared for the RPCJUA by Eqecat on June 23, 1997, based on the RPCJUA's April 30, 1997 exposures:

The RPCJUA's annual expected hurricane loss is $215 million.
The RPCJUA's 100-year probable maximum loss is $4.1 billion.

The projections would probably be lower based on the RPCJUA's current exposures. However, the projections would have been higher if Eqecat had included "time-element" coverages such as additional living expenses (the RPCJUA supplied no information on these coverages), or if it had included "demand surge," the expected increase in the cost of goods and services after a major disaster (the RPCJUA instructed Eqecat to exclude demand surge).

Table 7 shows the RPCJUA's expected annual hurricane loss for the 10 counties with the highest expected losses.

Table 7: Expected annual RPCJUA hurricane losses by county*

County Expected annual RPCJUA hurricane loss Expected annual RPCJUA hurricane loss in
the county as a percentage of total expected
annual RPCJUA hurricane loss

Dade $72.8 million 33.8%

Palm Beach $59.9 million 27.9%

Broward $48.4 million 25.5%

Pinellas $7.9 million 3.7%

Martin $4.6 million 2.1%

Lee $2.6 million 1.2%

Manatee $2.6 million 1.2%

Brevard $2.6 million 1.2%

Hillsborough $2.1 million 1.0%

St. Lucie $2.0 million 0.9%

10-county total $205.5 million 95.6%

*Source: Eqecat Catastrophe Management and RPCJUA

Table 7 indicates that the RPCJUA's hurricane loss potential is concentrated in Dade, Broward, and Palm Beach Counties, which are together responsible for 87.2% of the RPCJUA's annual expected hurricane loss of $215 million.

Residual property insurance markets:
Florida Windstorm Underwriting Association (FWUA)

The Florida Windstorm Underwriting Association was created in 1970 by subsection 627.351(2), Florida Statutes. The FWUA writes policies that cover only losses caused by windstorm, and writes these policies only in certain limited coastal areas. After Hurricane Andrew, the Insurance Department expanded FWUA eligibility to include Dade and Broward Counties east of I-95. Because of an amendment to FWUA eligibility in the 1996 property insurance law, properties east of I-95 in Palm Beach County and in coastal areas of Pasco County became eligible for the FWUA in 1997.

The FWUA provides personal lines and commercial lines property insurance policies (including both residential and non-residential policies) providing windstorm coverage to applicants who are unable to obtain coverage from an insurance company. In 1997, Chapter 97-55, Laws of Florida, (CS/SB 794) amended the FWUA law to specify that, as with the RPCJUA, an offer of coverage from a licensed insurer at Department of Insurance-approved rates disqualifies a policyholder or applicant from eligibility for FWUA coverage. Chapter 97-55 also froze any further geographic expansion of the FWUA until October 1, 1998.

The FWUA charges premiums for the coverage it provides, but, as with the RPCJUA, when the premiums and other resources of the FWUA are not sufficient to pay claims, the FWUA has the power to levy assessments on insurance companies and their policyholders to generate the revenues necessary to cover the deficit. The FWUA has the power to issue bonds and other debt instruments, and to pledge its premiums, assessments, and other resources to pay off the debt.

Beginning in 1999, FWUA premiums must be "reflective of" Department of Insurance-approved voluntary market rates for hurricane coverage. Prior to the adoption of Chapter 97-55 in 1997, the FWUA was subject only to the general ratemaking standards of section 627.062, F.S. Chapter 97-55 amended subparagraph 627.351(2)(b)5., F.S., to add the statement of legislative intent, identical to the intent language in the RPCJUA law, that FWUA rates be actuarially sound and not competitive with Department of Insurance-approved rates charged in the voluntary market. The 1997 amendments did not require that FWUA rates be matched to the rates of the top 20 insurance companies, but instead required that the FWUA plan of operation provide a mechanism to assure that, beginning in 1999, "the rates charged by the [FWUA] for each line of business are reflective of approved rates in the voluntary market for hurricane coverage for each line of business in the various areas eligible for [FWUA] coverage."

Subparagraph 627.351(2)(b)2., F.S., sets out the assessment procedures and limitations of the FWUA, which are generally the same as the procedures and limitations applicable to RPCJUA assessments. The differences are that the FWUA deficits and assessments are not divided into personal lines and commercial lines accounts, and that the policies subject to assessment include all property insurance policies, rather than being limited to residential property insurance policies. The current amount of premiums subject to FWUA assessment is $3.7 billion.

In 1995, the FWUA levied $117 million in assessments as a result of Hurricanes Erin and Opal. The 1995 assessment represented approximately 3.2% of the FWUA's assessment base.

The FWUA is currently experiencing rapid growth. The growth results in part from the recent expansions of eligibility, in part from the continuing unavailability of windstorm coverage in the most vulnerable areas of the state, and in part from insurers' implementation of "accelerated exposure reduction" plans. When section 627.7013, F.S., the moratorium on hurricane-related property insurance nonrenewals, was extended in 1996, insurers with extreme concentrations of risk were allowed to seek Department of Insurance approval to nonrenew a larger number of policies than would otherwise be allowed in one year, provided that the nonrenewals were limited to properties eligible for FWUA coverage (see s. 627.7013(2)(b)7., F.S.).

In the first eight months of 1997, the FWUA's policy count increased by 81,663 policies, or 28.9%, and its structure exposure increased by $15.7 billion, or 31.7%. In the same period in 1996, the policy count increased by 30,654 policies, or 13.3%, and structure exposure increased by $8.6 billion, or 23.7%.

Table 8 shows the county-by-county policy count and exposure of the FWUA as of August 31, 1997 and the net change in these items since December 31, 1996.

Table 8: FWUA policies in force and exposure as of August 31, 1997*

County Policies in force Exposure (in thousands)

As of 8/31/97 Net change since 12/31/96 % change since 12/31/96 As of 8/31/97 Net change since 12/31/96 % change since 12/31/96

Bay 7,230 1,317 22.27% $1,084,302 $178,459 19.70%

Brevard 4,762 834 21.23% $722,374 $88,511 13.98%

Broward 77,670 14,523 23.00% $12,442,581 $2,597,357 26.38%

Charlotte 1,469 117 8.65% $297,249 $14,676 6.18%

Collier 9,817 2,366 31.75% $2,670,354 $545,537 25.67%

Dade 87,926 13,039 17.41% $17,383,929 $2,890,733 19.95%

Duval 1,827 230 14.40% $316,147 $42,233 15.42%

Escambia 7,810 1,907 32.31% $1,238,998 $322,285 35.16%

Flagler 1,636 291 21.54% $218,907 $52,725 31.73%

Franklin 1,989 88 4.63% $283,509 $19,459 7.37%

Gulf 998 88 9.67% $118,229 $13,972 13.40

Hernando 638 137 27.35% $76,159 $18,373 31.79%

Indian River 3,390 1,003 42.02% $1,073,817 $330,878 44.54%

Lee 15,473 2,453 18.84% $2,941,754 $493,537 20.16%

Levy 438 21 5.04% $49,546 $3,210 6.93%

Manatee 3,913 347 9.73% $685,228 $70,271 11.42%

Monroe 30,356 1 0.00% $5,180,977 $197,975 3.97%

Nassau 1,290 229 21.58% $320,649 $43,201 15.57%

Okaloosa 3,397 826 32.13% $761,865 $133,851 21.31%

Palm Beach 22,556 22,556 N/A** $4,430,734 $4,430,734 N/A**

Pasco 2,478 2,478 N/A** $276,402 $276,402 N/A**

Pinellas 20,146 4,562 29.27% $3,523,302 $865,689 32.57%

St. Johns 2,914 359 14.05% $566,200 $39,509 7.50%

St. Lucie 3,433 369 12.04% $204,460 ($36,838) (15.27%)

Santa Rosa 1,195 355 42.26% $268,583 $92,869 52.85%

Sarasota 28,933 6,662 29.91% $4,989,621 $1,253,391 33.85%

Volusia 14,700 3,333 29.32% $1,819,725 $456,317 33.47%

Wakulla 445 46 11.53% $53,651 $8,450 18.69%

Walton 4,658 1,126 24.85% $1,192,041 $281,079 30.86%

TOTAL 364,487 81,663 28.87% $65,181,313 $15,704,945 31.74%

Dade-Broward-Palm Beach total 188,152 50,118 36.31% $34,257,244 $9,918,824 40.75%

Dade-Broward-Palm Beach share of state totals 51.62% 61.37% ------ 52.56% 63.16% ------

* Source: FWUA
** The geographical eligibility boundaries of the FWUA were expanded to include portions of Palm Beach and Pasco Counties in November, 1996. The first FWUA policies in these counties were written in 1997.

Although the FWUA grew rapidly in many parts of the state in the first eight months of 1997, growth was concentrated in Dade, Broward, and Palm Beach Counties, which were together responsible for 50,118 new FWUA policies, or 61.37% of the FWUA's growth.

The FWUA has relied on two modeling firms, Applied Insurance Research (AIR) and Risk Management Solutions (RMS), to calculate its probable maximum loss. Based on the FWUA's actual exposures as of June 30, 1997:

AIR determined that the 100-year probable maximum loss of the FWUA is $3.13 billion.
RMS determined that the 100-year probable maximum loss of the FWUA is $3.18 billion.

Probable maximum loss figures can be expected to grow as the FWUA grows. The FWUA projects that its 100-year probable maximum loss will be $4.20 billion as of December 31, 1997, and will be $4.91 billion as of December 31, 1998.

The FWUA used the AIR model to determine that its expected annual hurricane loss was $284 million as of August 31, 1997.

Table 9 summarizes the key differences between the RPCJUA and the FWUA.

Table 9: Comparison of key features of the RPCJUA and the FWUA

Issue RPCJUA FWUA

Coverage provided Full property insurance policies on residential properties. Windstorm-only policies on residential or non-residential property.

Geographic limitations Coverage available statewide, except that the RPCJUA is required to exclude windstorm coverage with respect to properties eligible for windstorm coverage from the FWUA. Limited geographic area: all of Monroe County; Dade, Broward, and Palm Beach Counties east of I-95 (east of US 1 in south Dade); cities of Pensacola, Gulf Breeze, Panama City Beach, Ormond Beach, Daytona Beach, South Daytona Beach, Sarasota, and Cedar Key; and a narrow coastal strip in most other coastal counties. Other areas may be added by the Department of Insurance upon petition, but 1997 legislation froze further expansion until 10/1/98.

Other eligibility limitations An offer from an admitted (licensed) insurer at Department of Insurance-approved rates disqualifies an applicant or policyholder from eligibility for RPCJUA coverage. Statute amended in 1997 to provide, as with the RPCJUA, that an offer from an admitted insurer at an approved rate disqualifies an applicant or policyholder from FWUA eligibility.

Rate standards Rates must be actuarially sound and not competitive with voluntary market, and must include a catastrophic factor reflecting the RPCJUA's actual exposures and lack of substantial surplus; rates must be at least as high as the highest Department of Insurance-approved rate for a comparable risk from one of the 20 largest insurers in the state. Beginning in 1999, FWUA rates must be reflective of Department of Insurance-approved rates in the voluntary market.

Assessment authority When a deficit occurs in one of the "accounts" (all personal lines residential policies, all commercial lines residential policies), the RPCJUA may levy regular and emergency assessments. Regular assessments are one-time assessments on insurers, which they can pass through to their policyholders; emergency assessments are levied on policyholders and collected by insurers for one or more years. 1997 legislation revised the assessment authority of the FWUA to make it the same as the RPCJUA, except that the FWUA does not separate deficits or assessments by "account."

Assessment limits Regular assessments: 10% of the statewide premium making up the assessment base (see "Assessment base," below) or 10% of the deficit, whichever is greater. Emergency assessments: 10% of the assessment base premiums or 10% of the deficit plus financing costs, whichever is greater, until the debt incurred by the RPCJUA is paid off. Same as RPCJUA limits, but applicable to a larger assessment base (see "Assessment base," below).

Assessment base Regular assessments: All voluntary market premiums in the same lines of business as are covered by the account in which the deficit arose. Emergency assessments: All voluntary market and residual market premiums in the same lines of business as are covered by the account in which the deficit arose. Regular assessments: All voluntary market property insurance premiums. Emergency assessments: All voluntary market and residual market property insurance premiums.

Governance A board consisting of the Insurance Consumer Advocate, 5 members designated by the insurance industry pursuant to the RPCJUA plan of operation, 5 consumer representatives appointed by the Insurance Commissioner, 2 representatives of the insurance industry appointed by the Insurance Commissioner. A board consisting of the Insurance Consumer Advocate, 1 consumer representative appointed by the Insurance Commissioner, and 12 members appointed by the insurance industry pursuant to the FWUA plan of operation. The Consumer Advocate and the consumer representative were added by 1997 legislation.

Florida Hurricane Catastrophe Fund (FHCF)

The Florida Hurricane Catastrophe Fund was created by section 215.555, Florida Statutes, in 1993 as Florida's "safety net" for a large hurricane. The Catastrophe Fund provides the equivalent of reinsurance to Florida residential insurers, funded through a combination of premiums paid by insurance companies and emergency assessments on all property and casualty insurance premiums except for workers' compensation premiums. The fund is administered by the State Board of Administration, which consists of the Governor, the Comptroller, and the Treasurer/Insurance Commissioner.

The Catastrophe Fund is required by contract with each residential property insurer to reimburse the insurer for a selected percentage (45%, 75%, or 90%) of the insurer's hurricane losses in excess of a specified amount, known as the insurer's "retention." The aggregate retention of all companies is currently $3.16 billion. This retention level assures that the fund will not pay any significant amount of claims in a minor hurricane, such as Hurricanes Erin and Opal, which struck Florida in 1995. In the aggregate, the fund is responsible for 84% of the insurance industry's covered losses in excess of the $3.16 billion retention. The fund collects approximately $450 million a year from insurers in exchange for this coverage.

In the event that the resources of the fund, including both the fund balance and debt instruments pledging emergency assessments, are insufficient to pay all claims against the fund, claims are to be paid on a pro-rata basis.

As of December 31, 1997, the balance in the fund will be $2 billion. The fund has estimated its bonding capacity (that is, the amount of bonds it would be able to issue given its current assessment base) at $5.5 billion.

Table 10 shows how the resources of the fund and of the insurance industry would be allocated after a hurricane causing $12 billion in insured residential losses in Florida. Of the $16 billion in insured losses from Hurricane Andrew, approximately $12 billion was attributable to the various forms of residential property coverage.

Table 10: The role of the Florida Hurricane Catastrophe Fund in a hypothetical hurricane comparable to Hurricane Andrew*

Amount of loss Payment sources

$12.1 billion Insurer reimbursement from Florida Hurricane Catastrophe Fund:
$7.5 billion
(composed of $2 billion fund balance, plus $5.5 billion in revenue bond proceeds)
$7.5 billion is 84% of the $8.9 billion loss in excess of industry retention.
Based on aggregate fund coverage level of 84% (weighted average of various coverage levels selected by insurers)
Insurance industry "co-pay":
$1.4 billion

$3.2 billion

Insurance industry retention: $3.2 billion

*Based on data provided by FHCF

Insurance companies and the residual market entities would be responsible, in the aggregate for the first $3.2 billion of covered residential hurricane losses. With respect to losses in excess of that threshold, insurance companies and the residual market would, in the aggregate, be responsible for $1.4 billion (or 16%) of the covered losses, and the fund would be responsible for $7.5 billion (or 84%) of the covered losses. The split between the fund's 84% and the industry's 16% is based on the Catastrophe Fund coverage levels (90%, 75%, or 45%) selected by insurers, which aggregate to 84%.

This analysis leaves out losses attributable to Additional Living Expense (ALE) coverages, which are not "losses" for purposes of Catastrophe Fund reimbursement (see s. 215.555(2)(d), F.S.). ALE coverage compensates policyholders for housing costs while they are waiting for their homes to be repaired, rebuilt, or replaced. According to memoranda from the Catastrophe Fund and from Applied Insurance Research, ALE losses represent approximately 5 percent of total residential losses, on average. The percentage of the total loss attributable to ALE will vary with the severity of the event; if overall losses are small, homeowners will be able to get back in their homes much sooner than if losses are large. If ALE losses were added to the losses shown in the example in Table 10 (assuming that ALE losses were 5 percent of the total residential loss), the total residential loss would be $12.7 billion, and the insurance industry share of that loss would be $5.2 billion.

Premiums paid to the Catastrophe Fund are based on four factors: rating territory, construction type, deductible amounts, and type of business (i.e., personal lines residential, commercial lines residential, or mobile home). The fund has rated each zip code in the state for its potential hurricane loss and has converted those ratings into 25 rating territories.

Table 11 shows examples of Catastrophe Fund premiums in various rating territories. These examples show the premium an insurer would pay for each $100,000 dwelling in the territory covered by a policy that has a $500 hurricane deductible. Premiums are shown for both wood-frame houses and for masonry houses. Catastrophe Fund premiums are lower for policies that have higher deductibles.

Table 11: Examples of premiums paid by insurers to the Florida Hurricane Catastrophe Fund for each $100,000 of personal lines residential exposure with a $500 deductible*

Catastrophe Fund Rating Territory Location (examples) Premium for frame construction Premium for joisted masonry construction

2 Jacksonville, Tallahassee $12 $9

3 Jacksonville Beach, Gainesville $17 $13

6 Clearwater, Pensacola $45 $35

7 Orlando, Cocoa Beach, Apalachicola $52 $40

8 Bradenton, Daytona Beach $57 $45

9 Tampa, Lakeland $65 $51

11 Vero Beach, St. Petersburg Beach $75 $59

20 West Palm Beach, Sanibel $180 $146

21 Fort Lauderdale, Ft. Myers Beach $190 $156

24 Miami Beach, Key West $229 $184

*Based on data provided by FHCF

The Catastrophe Fund premium structure is designed to reflect the relative likelihood that various geographic areas will sustain economic loss from major hurricanes. The premium variation among the rating territories is indicative of the variation in hurricane vulnerability within Florida, but the Catastrophe Fund premium plan should be used only as a general guide. The premium plan does not reflect losses that will not be covered by the Catastrophe Fund. Because the insurance industry, in the aggregate, is required to retain liability for the first $3.16 billion of insured residential hurricane losses, Catastrophe Fund premiums do not account for the various territories' likelihood to sustain losses below this amount.

Under subsection 215.555(6), Florida Statutes, when the State Board of Administration determines that a hurricane has occurred and the balance in the Catastrophe Fund will not be sufficient to meet all of the fund's obligations under its contracts with insurers, the board is authorized to provide for the issuance of revenue bonds on behalf of the fund and the levy of emergency assessments to support those bonds. All property and casualty insurance policies except for workers's compensation policies are subject to assessment; that is, Catastrophe Fund assessments apply not only property insurance policies, but also to automobile, commercial multi-peril and liability, medical malpractice, professional malpractice, and other policies classified as "property and casualty." The maximum amount of the emergency assessment is 2% of premium, except that the maximum increases to 4% of premium after the Governor declares a state of emergency.

Part III: Subsidies in Residual Market Assessments

The recoupment process

An insurer may use a uniform percentage surcharge on premiums to recover ("recoup") from its policyholders the regular assessments that it has paid to the RPCJUA or FWUA. Section 627.3512, F.S., was enacted in 1995 to clarify the recoupment process. Under that provision, regular assessments paid to the RPCJUA and FWUA, as well as assessments of two other residual market entities, the Florida [Automobile] Joint Underwriting Association and the Florida Medical Malpractice Joint Underwriting Association, are subject to recoupment by means of a premium surcharge.

: Section 627.3512, F.S., requires in part that:; The recoupment shall be made by applying a separate assessment factor on policies of the same line or type as were considered by the residual markets in determining the assessment liability of the insurer or insurer group.

Residual market assessments are recouped by means of a percentage surcharge, referred to in the statute as a "separate assessment factor." As applied by the Department of Insurance and insurance companies, the premiums in all lines relevant to the assessment are subject to an equal percentage surcharge. For example, when the RPCJUA personal lines account sustains a deficit, all personal lines residential premiums are used in determining the market shares that establish each insurer's share of the total personal lines assessment. An insurer's only option to recoup the personal lines assessment would be to apply a specific percentage surcharge to all personal lines residential premiums. It would not, for example, be able to impose different percentage surcharges on fire policies and multi-peril homeowners' policies, nor would it be able to impose any surcharge on commercial lines residential policies or automobile policies.

The insurer must give the Department of Insurance notice of its "assessment factor" and an explanation of how it will be applied at least 15 days before collecting any surcharges. The department's review of the recoupment filing is limited to verification of the insurer's arithmetic calculations.

Emergency assessments are levied as uniform percentages of premium and collected from policyholders by insurance companies, as distinguished from regular assessments, which are levied on insurance companies as dollar amounts and then recouped from policyholders.

Residual market assessments as subsidies

In some sense, all residual market assessments result in some consumers paying the insurance costs of other consumers. An assessment is a mandate that some insurers and their policyholders pay an additional amount -- over and above the amount that insurers and the Department of Insurance have determined is the appropriate cost of insuring their own homes -- to cover the insurance costs of policyholders insured by the RPCJUA and FWUA. These payments amount to subsidies under a standard dictionary definition of "subsidy": financial assistance given by one person or government to another (American Heritage Dictionary).

The statutory structure of both residual market entities (see pages 17-28, above) makes it clear that, at some point, all voluntary market consumers will subsidize the insurance costs of residual market consumers. The point at which voluntary market consumers were meant to begin subsidizing the insurance costs of residual market consumers is not as clear. However, there are strong statutory indications that subsidies should be imposed not in the case of expected residual market losses, but only in the case of extraordinary losses (see pages 17-19 and 24, above).

A particular policyholder's share of a residual market assessment is determined by applying a percentage surcharge to the policyholder's property insurance premium. In the case of regular assessments, the percentage is the same for all of an insurer's policyholders, but may vary from one insurer to another. Emergency assessments are collected as a uniform percentage surcharge on all policyholders' premiums, regardless of which insurer collects the surcharge.

Any significant deficit of the RPCJUA, and any deficit of the FWUA, will be the result of hurricane losses. When assessments are recouped or collected as percentage surcharges on the entire property insurance premium, rather than only the hurricane portion of the premium, a consumer's share of the total subsidy does not reflect that consumer's vulnerability to hurricane losses.

Homeowners' insurance premiums vary widely from one area of the state to another, and from one insurer to another. Table 12 shows homeowners' premium examples by county, based on the most recent officially-reviewed compilation of homeowners' premium examples. There are significant problems with the data in the table, but the table remains useful for broad comparisons. The table shows premium examples for a standard HO-3 homeowners' policy covering a $75,000 frame house with a hurricane deductible of either $250 or $500. Where an insurer used more than one rating territory in a county (commonly, coastal areas and inland areas will be in different rating territories), an unweighted, arithmetic average was used to determine the premium for the county. The problems with the data are: although the premium examples were taken from the May, 1997, rate filing of the RPCJUA, some premium increases have taken effect since the information was compiled; the premium for a $75,000 house may be misleading in areas where the median property value is well in excess of that amount or with respect to insurers that do not ordinarily insure properties of that value; the use of low deductibles may be misleading as the market moves toward higher deductibles, especially with respect to higher-value properties; and the use of unweighted averages within a county could give misleading results, especially in counties where both population and property values are concentrated in coastal areas.

Table 12 shows, for the 15 largest counties, the highest, lowest, and average (weighted by market share) Department of Insurance-approved premium for a standard homeowners' policy covering a $75,000 frame house for the top 20 homeowners' insurers in Florida. A range is shown in each of these categories to reflect the difference between data based on $500 deductibles and data based on $250 deductibles.

Table 12: Premium examples for a homeowners' policy covering a $75,000 frame house, 15 largest counties*

County Low High Average

Dade $618-$668 $1,312-$1,990 $1,004-$1,103

Broward $455-$599 $1,063-$1,771 $852-$869

Palm Beach $508-$510 $1,012-$1,397 $778-$779

Pinellas $325-$351 $736-$989 $458-$471

Hillsborough $349-$365 $720-$1,179 $500-$529

Orange $286-$287 $564-$966 $429-$435

Duval $310-$332 $598-$716 $373-$393

Polk $286-$379 $625-$1,051 $471-$488

Brevard $349-$354 $671-$1,026 $461-$489

Volusia $339-$349 $620-$995 $380-$446

Lee $406-$452 $795-$1018 $488-$553

Seminole $312-$384 $574-$919 $439-$447

Pasco $373-$399 $699-$795 $463-$469

Sarasota $388-$394 $746-$1,084 $465-$503

Escambia $370-$395 $735-$842 $486-$490

*Premium data taken from Exhibit VII of the RPCJUA's May, 1997 rate filing.

The impact of various premium levels on the level of a particular policyholder's subsidy of other policyholders can be seen from a simple hypothetical example. For purposes of this hypothetical, assume that a hurricane in Dade County has caused the RPCJUA to sustain a $250 million deficit, and that a one-time 10% assessment has been imposed to defray the deficit.

A Dade County homeowner who pays an "average" premium of $1,000 would be required to pay an assessment surcharge of $100, or 10% of the premium. An Orange County policyholder who pays an "average" premium of $430 would be required to pay an assessment surcharge of $43. The difference between the dollar amounts of the surcharges is a partial reflection of the difference between Dade County's vulnerability to hurricane losses and Orange County's vulnerability. However, the difference in surcharges is much less than the four-fold difference in hurricane vulnerability between Dade County and Orange County (see Table 11 on page 32, above).

The wide variation among Department of Insurance-approved premiums in any particular area of the state assures that at least some homeowners in less-vulnerable areas will pay more in surcharges than at least some owners of comparable homes in more-vulnerable areas. For example, if the Orange County homeowner in the example above were insured by a company that charged a "high" premium for Orange County -- $700, and the Dade County homeowner were insured by a company that charged a "low" premium for Dade County -- $650, the 10% assessment would result in the Orange County homeowner paying a larger surcharge than the Dade County homeowner. The difference in that case would be entirely attributable to the range of Department of Insurance-approved premiums among companies; a $75,000 Dade County home covered by a policy with a $650 premium is not necessarily any more hurricane-resistant or less vulnerable to hurricanes than a $75,000 Dade County home covered by a policy with a $1,000 premium. Similarly, the difference in premiums between the two Orange County homes does not necessarily reflect any difference in hurricane vulnerability.

Under the current assessment system, voluntary market policyholders will at some point pay some of the insurance costs of residual market policyholders, who are concentrated in the most vulnerable areas of the state. Because assessment surcharges are uniform percentages of the total property premium, the assessment system does not allow the amount of surcharges to reflect fully any geographic differences in hurricane vulnerability. Because of the wide range of Department of Insurance-approved premiums in the voluntary market, at least some homeowners in less-vulnerable areas will pay higher surcharges than owners of comparable homes in more-vulnerable areas.

Future assessment scenarios

Both the RPCJUA and the FWUA have financing plans in place that they indicate will enable them to pay claims from any hurricane up to the 100-year storm. These financial arrangements do not transfer risks from the residual market to Wall Street; instead, they merely defer the costs of a hurricane until after the storm occurs. The existence of the financing plans does not change the basic residual market funding issue: who will pay, and how much?

This report presents estimates of assessments likely to be generated by the residual market in a variety of scenarios. The scenarios are based on a number of assumptions, but because of the range and nature of the assumptions, the estimates are not intended to be precise projections. The estimates are meant to provide "ballpark" indications of the impact on the public of several foreseeable hurricane losses.

These are the assumptions that are used in the residual market assessment scenarios:

Assessment base: For the RPCJUA, assume that the voluntary market statewide aggregate premium for personal lines residential insurance is $2.5 billion, and the aggregate personal lines residential premium in the residual market is 0.75 billion. For purposes of FWUA assessments, the FWUA assumes a 1998 assessment base of $4.6 billion.
Exposures: Assume that the exposures of the RPCJUA and FWUA are the same as the mid-1997 exposures on which their most recent actuarial analyses were based. Neither of the residual market entities has a stable portfolio of exposures. As indicated above (see pages 25-28), the FWUA is growing rapidly in the most vulnerable areas of the state. For most of 1997, the number of policies removed from the RPCJUA under depopulation programs was roughly equal to the number of new policies issued by the RPCJUA.
Premiums and other revenues. Assume that the RPCJUA and FWUA are collecting premiums at mid-1997 levels, to match the mid-1997 exposures used for the projections.
Amounts available to pay hurricane losses without assessments. According to discussions with the Chief Executive Officer and Chief Financial Officer of the RPCJUA, it is reasonable to assume that the RPCJUA will enter 1998 with a surplus of approximately $100 million, and that this is the amount that would be available to pay RPCJUA claims before there was a need to assess, regardless of the size of the hurricane. The FWUA provided separate calculations of potential assessments for various scenarios.
Treatment of reinsurance and Catastrophe Fund reimbursements. Both of the residual market entities would receive reimbursement from the Florida Hurricane Catastrophe Fund after very large hurricane losses. In 1996, the RPCJUA's Catastrophe Fund retention was $346 million, and the FWUA's retention was $164 million. The 1997 figures, which will be finalized based on September 30, 1997, exposures, are expected to be considerably different from the 1996 retentions. In the case of the RPCJUA, Catastrophe Fund reimbursements are not considered when the amount of the deficit is calculated, and the assessment assumptions in this report reflect this approach to calculating the deficit. The FWUA applies reinsurance and Catastrophe Fund reimbursements to its deficit calculations; this report relies on their projections. (The RPCJUA's method may overstate the deficit, and thereby overstate the assessment authority of the RPCJUA, by treating revenues received from the Catastrophe Fund or other reinsurance as if they were not revenues.)

Table 13 identifies the loss projections used in the hurricane scenarios. Eqecat performed the RPCJUA's loss projections based on April 30, 1997 exposures (see pages 22-23, above, for a discussion of the limitations of these loss projections). The FWUA's loss projections were performed by AIR and were based on August 31, 1997 exposures. The scenarios are based on five levels of hurricane loss: the expected annual loss, and the 5, 10, 20, and 50 year probable maximum losses.

Table 13: Hurricane loss projections used in residual market assessment scenarios (in millions)*

Loss RPCJUA FWUA

Expected (average) annual hurricane loss $215 $284

5 year (20% probability) probable maximum loss $74 $375

10 year (10% probability) probable maximum loss $339 $669

20 year (5% probability) probable maximum loss $896 $1,575

50 year (2% probability) probable maximum loss $2,879 $3,031

*Source: RPCJUA data from Eqecat, FWUA data from AIR

Loss scenario 1: Expected annual loss

The expected annual hurricane loss of the RPCJUA is $215 million; the expected annual hurricane loss of the FWUA is $284 million.

In the event of a $215 million loss, the RPCJUA would sustain a deficit of $115 million. A $115 million deficit would be covered through a one-time regular assessment. The $115 million assessment would result in assessment surcharges to each voluntary market personal lines residential policyholder of approximately 4.6% of premium.

In the event of a $284 million loss, the FWUA would sustain a deficit of $149 million, which would be covered through a one-time regular assessment. The $149 million assessment would result in assessment surcharges to each property insurance policyholder of approximately 3.2% of premium.

Loss scenario 2: 5-year probable maximum loss

The 5-year probable maximum loss of the RPCJUA is $74 million. The RPCJUA would be able to sustain a $74 million hurricane loss without incurring a deficit.

The 5-year probable maximum loss of the FWUA is $375 million. A $375 million loss would result in a one-time deficit assessment of $241 million. A $241 million assessment would produce surcharges of approximately 5.2% of all property insurance premiums.

Loss scenario 3: 10-year probable maximum loss

The 10-year probable maximum loss of the RPCJUA is $339 million; the 10-year probable maximum loss of the FWUA is $669 million.

For the RPCJUA, a $339 million loss would result in a one-time deficit assessment of $239 million, or about 9.6% of all personal lines residential premiums. This is nearly the largest loss that the RPCJUA could sustain without resorting to multi-year assessments.

For the FWUA, a $669 million hurricane loss would result in deficit assessments of $450 million, after application of $92 million in Catastrophe Fund recoveries. This assessment amount is about 9.7% of all property insurance premiums. As with the RPCJUA, the 10-year probable maximum loss is nearly the largest loss that the FWUA could sustain without resorting to multi-year assessments.

Loss scenario 4: 20-year probable maximum loss

The 20-year probable maximum loss of the RPCJUA is $896 million; the 20-year probable maximum loss of the FWUA is $1,575 million.

In the event of a hurricane loss of $896 million, the RPCJUA would sustain a deficit of $796 million, plus financing costs. This deficit would result in a one-time regular assessment of $250 million, or 10% of personal lines residential property premiums, and a series of multi-year emergency assessments. If the RPCJUA levied the maximum amount of emergency assessments (which would in this case be $325 million, or 10% of all voluntary market and residual market residential premiums), two years of emergency assessments would be sufficient to cover the deficit and financing costs. However, the RPCJUA could elect to pay back its debt using a lower assessment level for a greater number of years.

The FWUA's 20-year probable maximum loss would result in a deficit of $702 million, after considering $703 million in Catastrophe Fund recoveries and $58 million in other reinsurance recoveries. In the event of a $702 million deficit, the FWUA would use a one-time 10% regular assessment on all property insurance premiums, followed by a one-time 5.2% emergency assessment on all property insurance premiums.

Loss scenario 5: 50-year probable maximum loss

The 50-year probable maximum loss of the RPCJUA is $2,879 million; the 50-year probable maximum loss of the FWUA is $3,031 million.

In the event of a hurricane loss of $2,879 million, the RPCJUA would sustain a deficit of $2,779 million, plus financing costs. This deficit would result in a one-time regular assessment of $278 million, or 10% of the deficit, and a series of multi-year emergency assessments. If the RPCJUA levied the maximum amount of emergency assessments (which would in this case be $325 million, or 10% of all voluntary market and residual market residential premiums), up to 10 years of emergency assessments at the maximum level would probably be needed to cover the deficit and financing costs.

The FWUA's 50-year probable maximum loss would result in a deficit of $1,648 million, after considering $1 billion in Catastrophe Fund recoveries and $300 million in other reinsurance recoveries. In the event of a $1,648 million deficit, the FWUA would use a one-time 10% regular assessment on all property insurance premiums, followed by 3 years of 8.6% emergency assessments on all property insurance premiums.

Table 14 summarizes the likely outcomes of the various loss scenarios.

Table 14: Summary of deficit assessments in 5 hurricane loss scenarios*

Scenario RPCJUA assessments FWUA assessments

Regular Emergency Regular Emergency

Expected annual loss 4.6% of voluntary market personal lines residential premiums, one-time. None. 3.2% of all property insurance premiums, one-time. None.

5-year probable maximum loss None. None. 5.2% of all property insurance premiums, one time. None.

10-year probable maximum loss 9.6% of voluntary market personal lines residential premiums, one time. None. 9.7% of all property insurance premiums, one time. None.

20-year probable maximum loss 10% of voluntary market personal lines residential premiums, one time. 10% of all personal lines residential premiums, two years, or lower assessments for a longer time period. 10% of all property insurance premiums, one time. 5.2% of all property insurance premiums, one time.

50-year probable maximum loss 11.1% of voluntary market personal lines residential premiums, one time. 10% of all personal lines residential premiums, 10 years. 10% of all property insurance premiums, one time. 8.6% of all property insurance premiums, three years, or lower assessments for a longer time period.

* Sources: RPCJUA resources based on discussions with RPCJUA officials cited above, RPCJUA loss projections from Eqecat, FWUA resources and projected assessments from FWUA response to questionnaire, FWUA projected losses from AIR.

According to these projections, the highest combined assessment that would be surcharged to a homeowners' policy after a 50-year storm would be 21.1% of the premium, followed by several years of 18.6% surcharges, and several years of 10% surcharges after that. A hurricane of that magnitude could also be expected to result in emergency assessments from the Florida Hurricane Catastrophe Fund (see Part IV, below) and the Florida Insurance Guaranty Association.

Part IV: Subsidies and Catastrophe Fund "Recoupment"

Statutory treatment of Catastrophe Fund premiums

The premiums an insurance company pays to the Catastrophe Fund are functionally the same as the premiums an insurance company pays to a private sector reinsurer. In both cases, the insurer transfers a portion of its risk to another entity and pays a premium to the risk-assuming entity.

All admitted insurance companies writing residential property insurance in Florida are required by law to participate in the Catastrophe Fund, just as they are required by law to participate (in the sense of being subject to assessments) in the residual market. Catastrophe Fund premiums, however, do not resemble residual market assessments. An insurer does not receive any direct benefit in exchange for paying residual market assessments, but an insurer receives something of value -- a binding promise to reimburse a portion of the insurer's hurricane losses -- in exchange for paying Catastrophe Fund premiums. Unlike residual market assessments, which are levied on insurers in proportion to their statewide market share of premiums written, Catastrophe Fund premiums are based on actuarial analyses of each insurer's portfolio of exposures.

The distinction between residual market assessments and Catastrophe Fund premiums was first addressed by the Legislature in the law that established the Catastrophe Fund in 1993, Chapter 93-409, Laws of Florida. The provisions on calculation of Catastrophe Fund premiums required the insurer to pay "an actuarially indicated premium" on a zip code-by-zip code basis for the Catastrophe Fund reimbursement contract. These provisions are currently part of subsection 215.555(5), F.S.

The Catastrophe Fund law addresses the manner in which Catastrophe Fund premium payments are to be treated by insurance companies. In language that has not changed since the law was adopted, paragraph 215.555(5)(d), F.S., provides:

(d) All premiums paid to the fund under reimbursement contracts shall be treated as premium for approved reinsurance for all accounting and regulatory purposes.

As stated in the final House Insurance Committee staff analysis of the original Catastrophe Fund law, "This provision recognizes that reimbursement from the fund is the functional equivalent of reinsurance."

Reinsurance costs or the costs of access to capital necessary to pay claims are included in a portion of each insurer's rate base known as the "catastrophe load." Usually an insurer's rating plan includes a catastrophe load based on the insurer's exposure. The catastrophe load is intended to reflect the annualized costs of catastrophes, the cost of access to the capital necessary to cover a foreseeable worst-case scenario (usually, the cost of reinsurance or financing to cover a company's 100-year probable maximum loss), and related expenses.

The adjustments needed to incorporate Catastrophe Fund premiums into a catastrophe load vary among insurers. For companies that had adequate or near-adequate catastrophe loads in their rate plans when the Catastrophe Fund began operating, little or no rate adjustment was necessary to reflect Catastrophe Fund costs, because the catastrophe load already reflected the full cost of insuring against catastrophes. To the extent of a prior catastrophe load's inadequacy, the Catastrophe Fund imposes new costs on an insurer, but the issue of catastrophe load inadequacy would have been present with or without the Catastrophe Fund. In fact, without the tax-exempt, non-profit Catastrophe Fund, an insurer would have to pay even more to the private reinsurance market or the financial markets to enable itself to pay policyholders' claims after a disaster.

Neither the Catastrophe Fund law nor the property insurance rating law explicitly allows or prohibits the use of a uniform percentage surcharge as a means of "recouping" an insurer's Catastrophe Fund costs. Some insurers have argued that the use of the word "recoup," which implies a retrospective review of actual costs, rather than a prospective inclusion of expected costs, indicates that the Legislature intended to authorize uniform percentage surcharges as a method of recoupment. Although the text of the law might support such a construction, there is nothing in the legislative history of the provision indicating that this was the legislative intent.

Catastrophe Fund costs are statutorily treated as part of the premium charged by the insurer. Paragraph 215.555(5)(d), F.S., requires that, for all regulatory and accounting purposes, Catastrophe Fund premiums be treated the same as reinsurance costs. Subsection 627.062(5), F.S., states that "the insurer may fully recoup in its property insurance premiums" any reimbursement premiums paid to the Catastrophe Fund.

The rating law provision quoted above addresses the amount of Catastrophe Fund costs an insurer is entitled to build into its rate filing, but does not explicitly address the method by which the insurer is to recover that amount from its policyholders. Since Catastrophe Fund premium recoupments are part of the premium charged by insurers, and subject to the rating law, it is possible that a percentage surcharge based on a home's total insurance premium (rather than only the hurricane portion of that premium) might be "unfairly discriminatory" in violation of the rating law (see page 12, above).

The impact of uniform percentage surcharges

The Catastrophe Fund collects approximately $450 million a year in reimbursement premiums from residential property insurers, including both the residual market and the voluntary market. Assuming total Florida residential property premiums of $3.5 billion a year, Catastrophe Fund costs amount to 13% of the total residential property premium.

Catastrophe Fund costs as a percentage of premium written will vary among insurers. Insurers with identical Catastrophe Fund costs (that is, insurers that have identical portfolios of exposures and that have selected identical Catastrophe Fund coverage levels) will not necessarily be authorized by the Department of Insurance to collect identical amounts of premium from their policyholders.

To the extent that a Catastrophe Fund "recoupment" results in an insurer collecting from one homeowner more than what the insurer paid to cover that homeowner's risk, and collecting from another homeowner less than what the insurer paid to cover the other homeowner's risk, one homeowner subsidizes the other homeowner (see the definition of "subsidy" on page 36).

The following hypothetical example illustrates how percentage surcharges can operate as a subsidy system. Assume that an insurer covers five $100,000 wood frame houses, located in Ft. Lauderdale, Lakeland, Orlando, Pensacola, and Jacksonville. Consistent with the averages shown in Table 12 on page 38, above, the insurer charges base premiums of $1,100 in Ft. Lauderdale, $650 in Lakeland and Pensacola, $600 in Orlando, and $450 in Jacksonville. The insurer "recoups" its Catastrophe Fund costs through a surcharge of 12% on all residential property premiums.

In this example, the insurer would have paid the Catastrophe Fund $190 for the exposure represented by the Ft. Lauderdale home, $65 for the exposure represented by the Lakeland home, $52 for the exposure represented by the Orlando home, $45 for the exposure represented by the Pensacola home, and $12 for the exposure represented by the Jacksonville home (see Table 11 on page 32, above). Applying the 12% surcharge, the insurer would collect $132 from the Ft. Lauderdale homeowner, $78 from the Lakeland homeowner and the Pensacola homeowner, $72 from the Orlando homeowner, and $54 from the Jacksonville homeowner.

Table 15 summarizes the hypothetical example.

Table 15: Summary of hypothetical example of uniform percentage surcharges for Catastrophe Fund premium recoupment*

Location of $100,000 frame house Homeowner's insurance premium Premium paid to Catastrophe Fund for exposure represented by $100,000 frame house in specified location Amount of surcharge on homeowner's premium used to "recoup" premium paid to Catastrophe Fund Amount of subsidy (paid) or received

Jacksonville $450 $12 $54 ($42)

Pensacola $650 $45 $78 ($33)

Orlando $600 $52 $72 ($20)

Lakeland $650 $65 $78 ($13)

Ft. Lauderdale $1,100 $190 $132 $58

*Sources: see discussion above. Assumptions: $100,000 wood frame house, standard homeowner's policy with $500 deductible, insurer uses 12% premium surcharge to "recoup" premiums paid to Catastrophe Fund.

Catastrophe Fund "recoupment" surcharges in the marketplace

There is no consistent insurance industry method for including or "recouping" Catastrophe Fund premium costs in the premiums charged to property owners. Some insurers are currently using uniform percentage surcharges of the total homeowners' insurance premium, while others build Catastrophe Fund costs into the catastrophe load of their rate base. In a draft rule that was proposed in 1996 and subsequently withdrawn (draft rule 4-170.016), the Department of Insurance would have allowed uniform percentage recoupment surcharges and would have provided for expedited review of filings using this method of recoupment. There are currently no rules in effect governing methods of Catastrophe Fund recoupment.

Both the RPCJUA and the FWUA currently use uniform percentage surcharges to recoup their Catastrophe Fund premiums. The FWUA's currently-pending rate filing would treat Catastrophe Fund premiums as part of the base rate, rather than recoup them as surcharges. Some large insurance companies, including State Farm and USAA, use percentage surcharges for Catastrophe Fund recoupment; other large insurance companies, including Allstate, Nationwide, and Florida Farm Bureau, do not use percentage surcharges for Catastrophe Fund recoupment.

Catastrophe Fund emergency assessments

In addition to charging premiums for the reinsurance coverage it provides, the Catastrophe Fund is able to rely on assessments of insurance premiums. As described above (see page 33), the Catastrophe Fund has the power to levy assessments on all property and casualty insurance premiums other than workers' compensation premiums when the fund's other resources are insufficient to meet the fund's obligations. The emergency assessments are capped at 4% of premium.

Catastrophe Fund emergency assessments result in subsidies of the costs of one kind of insurance (residential property coverage) by persons who pay premiums for other kinds of insurance, including private passenger and commercial motor vehicle insurance, commercial multiperil insurance, and all forms of liability and malpractice insurance. As with residual market assessments, Catastrophe Fund emergency assessments also result in low-risk homeowners subsidizing high-risk homeowners, since a Catastrophe Fund emergency assessment would be collected as a uniform percentage surcharge on the entire multi-peril property insurance premium.

The Catastrophe Fund would not need to rely on assessments until the total insured residential loss exceeded approximately $6 billion, or about one-half of the total residential loss from Hurricane Andrew. In a storm comparable to Hurricane Andrew, the fund would probably need to levy the full amount of the assessment to generate the revenue that would support the 15-year, $5.5 billion bond issue that represents the officially-determined bonding capacity of the fund. (See Table 10 and the discussion on pages 30-32, above.) The amount of a loss that the fund could sustain without resorting to assessments increases by approximately $500 million (from premium revenues and investment income) each year, except for years in which hurricane losses deplete the balance of the fund.

Part V: Implications of Subsidized Insurance Costs

The arguments for subsidies

A number of justifications have been advanced over the years for subsidies in the Florida homeowners' insurance market. Supporters of subsidy systems generally make one or more of the following points:

Subsidies are needed to assure that RPCJUA and FWUA policies are "affordable." It is argued that property insurance premiums in the residual market are currently so high that the public will not tolerate any further premium increases. It is also argued that the substantial increases in premiums since Hurricane Andrew especially burden homeowners living on fixed incomes, and that many fixed-income homeowners will be forced out of their homes if their premiums continue to be "unaffordable."
Even if residual market premiums increase substantially, there will still be a need for subsidies after a hurricane as destructive as Andrew. The argument is that even if residual market premiums were much higher than current premiums, and even if the additional revenues were primarily used for additional reinsurance, there would still be storms so massive that some sort of subsidy would be needed to assure payment of all claims. The argument goes on to suggest that large numbers of unpaid claims would do severe harm to the state's economy, and that the economic harm could be more costly than the subsidies.
With respect to emergency assessments to support the Catastrophe Fund, it is argued that assessments of a wide range of property and casualty insurance consumers (including such areas as auto, liability, and malpractice) would not be required except in the case of unusually destructive hurricanes approaching the destructiveness of Hurricane Andrew. In those circumstances, property insurance losses could threaten the solvency of at least some companies writing both property insurance and other lines of property and casualty insurance. It is also argued that the 4% cap on Catastrophe Fund assessments assures that any one consumer's assessment is small in comparison to the broad public benefit.

The arguments against subsidies

The arguments for subsidies are balanced by a number of arguments against subsidies. These arguments typically make these points:
- Subsidies unfairly burden low-risk homeowners for the benefit of high-risk homeowners, even though high-risk homes are often more costly, and their owners more wealthy, than low-risk homes and their owners. In addition, because rural fire insurance premiums tend to be higher than urban fire insurance premiums, it is also argued that the subsidies place a further burden on low-risk rural homeowners.
- Subsidies encourage concentrated development in high-risk coastal areas by masking the market costs of coastal development.
- Subsidies remove an incentive for mitigation. It is argued that subsidies, by reducing insurance costs in high-risk areas, mask the full value of improvements that make a home more hurricane-resistant. Therefore, it is argued that subsidies diminish the value of incentives, such as insurance premium discounts, designed to encourage loss mitigation.
- Subsidies encourage further subsidies. The argument is that many homeowners choose to be insured in the residual market in situations where the (subsidized) residual market premium is lower than the (unsubsidized) voluntary market premium, and that "gatekeeper" restrictions intended to make the residual market a last resort are of limited efficacy. The result, it is argued, is that subsidized residual market premiums lead to growth of the residual market, which in turn increases the overall subsidy level.
Conclusion

This report raised two questions: to what extent are subsidies relied on to fund the residential property insurance system in Florida, and how are the subsidy costs distributed among Florida consumers?

The statutory structure of each of the three state-created property insurance entities, the RPCJUA, the FWUA, and the Catastrophe Fund, depends on subsidies (in the form of assessments of insurance companies and consumers) when the resources of the state-created entity are insufficient to meet its obligations. Although the Legislature might have contemplated that these subsidies occur only in the context of major catastrophic losses, the RPCJUA and FWUA appear to rely on assessments to cover losses caused by even minor catastrophic events.

The subsidy costs resulting from residual market hurricane losses are not distributed among Florida consumers in proportion to their hurricane vulnerability. In that sense, some owners of low-risk homes subsidize some owners of high-risk homes (for example, when an Orange County consumer pays an assessment surcharge resulting from a Dade County hurricane), but high-risk homeowners also subsidize other high-risk homeowners (for example, when a Dade County consumer pays an assessment surcharge resulting from a Dade County hurricane). Low-risk homeowners subsidize high-risk homeowners when insurers recoup their payments to the Catastrophe Fund (which are calculated according to hurricane vulnerability) by means of a uniform percentage surcharge, which does not reflect relative hurricane vulnerabilities.

Cogent arguments have been made in support of subsidies, to the effect that they are necessary to assure continued availability and affordability of insurance coverage, and cogent arguments have been made against subsidies, to the effect that they distort market incentives and unfairly burden some consumers for the benefit of other consumers.

Table 1: Country-wide premiums (in millions ) and combined ratios for selected lines of property and casualty insurance, 1990-1995*
	1990	1991	1992	1993	1994	1995
Private passenger auto liability
Premiums written	$47,831	$51,200	$55,473	$59,273	$61,952	$65,337
Combined ratio**	117.7	113.3	109.3	107.7	104.9	101.8
Private passenger auto comprehensive and collision
Premiums written	$30,562	$31,555	$32,898	$34,102	$34,861	$36,639
Combined ratio**	93.9	88.5	87.7	88.9	92.6	97.2
Commercial multi-peril
Premiums written	$17,709	$17,032	$16,432	$17,308	$17,801	$18,845
Combined ratio**	107.6	110.8	126.7	115.5	118.7	112.4
Homeowners' multi-peril
Premiums written	$18,577	$19,303	$20,477	$21,546	$21,551	$23,987
Combined ratio**	112.5	117.3	158.0	113.1	118.0	112.3

Table 2: Total value of residential properties insured under policies providing windstorm coverage as of September 30, 1996 (in millions)*
County	Personal lines residential	Commercial lines residential	Mobile home	Total insured value	County percentage of statewide total insured value
Dade	$74,339	$9,037	$253	$83,629	10.6%
Broward	$71,980	$10,101	$733	$82,815	10.5%
Palm Beach	$71,771	$8,923	$557	$81,251	10.3%
Pinellas	$43,020	$5,534	$1,655	$50,209	6.4%
Orange	$39,597	$3,351	$693	$43,642	5.5%
Hillsborough	$39,373	$2,494	$1,095	$42,963	5.4%
All others	$349,697	$32,082	$22,610	$404,386	51.2%
State total	$689,777	$71,522	$27,596	$788,895

Table 3: Aggregate statewide property insurance premiums in Florida, 1993-1995 (in millions)*
Year	Personal lines residential			Commercial lines residential
	Voluntary market	Residual market	Total	Voluntary market	Residual market	Total
1993	$1,466	$116	$1,582	$722	$17	$739
1994	$1,670	$376	$1,976	$796	$51	$847
1995	$1,824	$449	$2,273	$890	$51	$941

Table 5: Share of insured value covered by the residual market, 6 largest counties, personal lines and commercial lines residential, as of September 30, 1996*
County	RPCJUA		FWUA		Voluntary market
	Exposure (in millions)	Percentage of county total	Exposure (in millions)	Percentage of county total	Exposure (in millions)	Percentage of county total
Dade	$13,953	16.7%	$13,340	15.9%	$56,336	67.4%
Broward	$12,256	14.8%	$8,738	10.6%	$61,821	74.6%
Palm Beach	$11,958	14.7%	0	0	$69,293	85.3%
Pinellas	$4,822	9.6%	$2,435	4.8%	$42,952	85.6%
Orange	$1,560	3.6%	0	0	$42,082	96.4%
Hillsborough	$3,561	8.3%	0	0	$39,402	91.7%
State total	$66,351	8.4%	$45,928	5.8%	$676,616	85.8%

Table 6: RPCJUA personal lines residential policy count and exposure as of September 30, 1997, excluding policies that do not provide windstorm coverage*
	Policies in force	Premium written (in millions)	Structure exposure (in millions)	Contents exposure (in millions)
Dade County	142,575	$152	$13,316	$5,571
Dade County as a percentage of state total	25.0%	32.5%	28.5%	27.8%
Broward County	98,320	$99	$9,681	$4,587
Broward County as a percentage of state total	17.3%	21.1%	20.7%	22.9%
Palm Beach County	64,335	$67	$7,354	$3,572
Palm Beach County as a percentage of state total	11.3%	14.3%	15.7%	17.8%
Pinellas County	50,473	$29	$3,362	$1,410
Pinellas County as a percentage of state total	8.9%	6.2%	7.2%	7.0%
Hillsborough County	30,376	$17	$2,102	$820
Hillsborough County as a percentage of state total	5.3%	3.6%	4.5%	4.1%
Top 5 counties	386,079	$364	$35,815	$15,959
Top 5 counties as a percentage of state total	67.8%	77.8%	76.7%	79.6%
State total	569,684	$468	$46,704	$20,039

Table 7: Expected annual RPCJUA hurricane losses by county*
County	Expected annual RPCJUA hurricane loss	Expected annual RPCJUA hurricane loss in the county as a percentage of total expected annual RPCJUA hurricane loss
Dade	$72.8 million	33.8%
Palm Beach	$59.9 million	27.9%
Broward	$48.4 million	25.5%
Pinellas	$7.9 million	3.7%
Martin	$4.6 million	2.1%
Lee	$2.6 million	1.2%
Manatee	$2.6 million	1.2%
Brevard	$2.6 million	1.2%
Hillsborough	$2.1 million	1.0%
St. Lucie	$2.0 million	0.9%
10-county total	$205.5 million	95.6%

Table 9: Comparison of key features of the RPCJUA and the FWUA
Issue	RPCJUA	FWUA
Coverage provided	Full property insurance policies on residential properties.	Windstorm-only policies on residential or non-residential property.
Geographic limitations	Coverage available statewide, except that the RPCJUA is required to exclude windstorm coverage with respect to properties eligible for windstorm coverage from the FWUA.	Limited geographic area: all of Monroe County; Dade, Broward, and Palm Beach Counties east of I-95 (east of US 1 in south Dade); cities of Pensacola, Gulf Breeze, Panama City Beach, Ormond Beach, Daytona Beach, South Daytona Beach, Sarasota, and Cedar Key; and a narrow coastal strip in most other coastal counties. Other areas may be added by the Department of Insurance upon petition, but 1997 legislation froze further expansion until 10/1/98.
Other eligibility limitations	An offer from an admitted (licensed) insurer at Department of Insurance-approved rates disqualifies an applicant or policyholder from eligibility for RPCJUA coverage.	Statute amended in 1997 to provide, as with the RPCJUA, that an offer from an admitted insurer at an approved rate disqualifies an applicant or policyholder from FWUA eligibility.
Rate standards	Rates must be actuarially sound and not competitive with voluntary market, and must include a catastrophic factor reflecting the RPCJUA's actual exposures and lack of substantial surplus; rates must be at least as high as the highest Department of Insurance-approved rate for a comparable risk from one of the 20 largest insurers in the state.	Beginning in 1999, FWUA rates must be reflective of Department of Insurance-approved rates in the voluntary market.
Assessment authority	When a deficit occurs in one of the "accounts" (all personal lines residential policies, all commercial lines residential policies), the RPCJUA may levy regular and emergency assessments. Regular assessments are one-time assessments on insurers, which they can pass through to their policyholders; emergency assessments are levied on policyholders and collected by insurers for one or more years.	1997 legislation revised the assessment authority of the FWUA to make it the same as the RPCJUA, except that the FWUA does not separate deficits or assessments by "account."
Assessment limits	Regular assessments: 10% of the statewide premium making up the assessment base (see "Assessment base," below) or 10% of the deficit, whichever is greater. Emergency assessments: 10% of the assessment base premiums or 10% of the deficit plus financing costs, whichever is greater, until the debt incurred by the RPCJUA is paid off.	Same as RPCJUA limits, but applicable to a larger assessment base (see "Assessment base," below).
Assessment base	Regular assessments: All voluntary market premiums in the same lines of business as are covered by the account in which the deficit arose. Emergency assessments: All voluntary market and residual market premiums in the same lines of business as are covered by the account in which the deficit arose.	Regular assessments: All voluntary market property insurance premiums. Emergency assessments: All voluntary market and residual market property insurance premiums.
Governance	A board consisting of the Insurance Consumer Advocate, 5 members designated by the insurance industry pursuant to the RPCJUA plan of operation, 5 consumer representatives appointed by the Insurance Commissioner, 2 representatives of the insurance industry appointed by the Insurance Commissioner.	A board consisting of the Insurance Consumer Advocate, 1 consumer representative appointed by the Insurance Commissioner, and 12 members appointed by the insurance industry pursuant to the FWUA plan of operation. The Consumer Advocate and the consumer representative were added by 1997 legislation.

Table 10: The role of the Florida Hurricane Catastrophe Fund in a hypothetical hurricane comparable to Hurricane Andrew*
Amount of loss	Payment sources
$12.1 billion	Insurer reimbursement from Florida Hurricane Catastrophe Fund: $7.5 billion (composed of $2 billion fund balance, plus $5.5 billion in revenue bond proceeds) $7.5 billion is 84% of the $8.9 billion loss in excess of industry retention. Based on aggregate fund coverage level of 84% (weighted average of various coverage levels selected by insurers)	Insurance industry "co-pay": $1.4 billion
$3.2 billion	Insurance industry retention: $3.2 billion

Table 11: Examples of premiums paid by insurers to the Florida Hurricane Catastrophe Fund for each $100,000 of personal lines residential exposure with a $500 deductible*
Catastrophe Fund Rating Territory	Location (examples)	Premium for frame construction	Premium for joisted masonry construction
2	Jacksonville, Tallahassee	$12	$9
3	Jacksonville Beach, Gainesville	$17	$13
6	Clearwater, Pensacola	$45	$35
7	Orlando, Cocoa Beach, Apalachicola	$52	$40
8	Bradenton, Daytona Beach	$57	$45
9	Tampa, Lakeland	$65	$51
11	Vero Beach, St. Petersburg Beach	$75	$59
20	West Palm Beach, Sanibel	$180	$146
21	Fort Lauderdale, Ft. Myers Beach	$190	$156
24	Miami Beach, Key West	$229	$184

Table 12: Premium examples for a homeowners' policy covering a $75,000 frame house, 15 largest counties*
County	Low	High	Average
Dade	$618-$668	$1,312-$1,990	$1,004-$1,103
Broward	$455-$599	$1,063-$1,771	$852-$869
Palm Beach	$508-$510	$1,012-$1,397	$778-$779
Pinellas	$325-$351	$736-$989	$458-$471
Hillsborough	$349-$365	$720-$1,179	$500-$529
Orange	$286-$287	$564-$966	$429-$435
Duval	$310-$332	$598-$716	$373-$393
Polk	$286-$379	$625-$1,051	$471-$488
Brevard	$349-$354	$671-$1,026	$461-$489
Volusia	$339-$349	$620-$995	$380-$446
Lee	$406-$452	$795-$1018	$488-$553
Seminole	$312-$384	$574-$919	$439-$447
Pasco	$373-$399	$699-$795	$463-$469
Sarasota	$388-$394	$746-$1,084	$465-$503
Escambia	$370-$395	$735-$842	$486-$490

Table 13: Hurricane loss projections used in residual market assessment scenarios (in millions)*
Loss	RPCJUA	FWUA
Expected (average) annual hurricane loss	$215	$284
5 year (20% probability) probable maximum loss	$74	$375
10 year (10% probability) probable maximum loss	$339	$669
20 year (5% probability) probable maximum loss	$896	$1,575
50 year (2% probability) probable maximum loss	$2,879	$3,031

Table 14: Summary of deficit assessments in 5 hurricane loss scenarios*
Scenario	RPCJUA assessments		FWUA assessments
	Regular	Emergency	Regular	Emergency
Expected annual loss	4.6% of voluntary market personal lines residential premiums, one-time.	None.	3.2% of all property insurance premiums, one-time.	None.
5-year probable maximum loss	None.	None.	5.2% of all property insurance premiums, one time.	None.
10-year probable maximum loss	9.6% of voluntary market personal lines residential premiums, one time.	None.	9.7% of all property insurance premiums, one time.	None.
20-year probable maximum loss	10% of voluntary market personal lines residential premiums, one time.	10% of all personal lines residential premiums, two years, or lower assessments for a longer time period.	10% of all property insurance premiums, one time.	5.2% of all property insurance premiums, one time.
50-year probable maximum loss	11.1% of voluntary market personal lines residential premiums, one time.	10% of all personal lines residential premiums, 10 years.	10% of all property insurance premiums, one time.	8.6% of all property insurance premiums, three years, or lower assessments for a longer time period.

Table 15: Summary of hypothetical example of uniform percentage surcharges for Catastrophe Fund premium recoupment*
Location of $100,000 frame house	Homeowner's insurance premium	Premium paid to Catastrophe Fund for exposure represented by $100,000 frame house in specified location	Amount of surcharge on homeowner's premium used to "recoup" premium paid to Catastrophe Fund	Amount of subsidy (paid) or received
Jacksonville	$450	$12	$54	($42)
Pensacola	$650	$45	$78	($33)
Orlando	$600	$52	$72	($20)
Lakeland	$650	$65	$78	($13)
Ft. Lauderdale	$1,100	$190	$132	$58