Document View

Pres Bush has made homeownership a central element of his housing policy. Through the American Dream Down Payment Act, the Homeowner's Tax Credit, and the Zero Downpayment Mortgage Initiative, Pres Bush is attempting to make homeownership easier both by reducing the down payment required and by increasing financing options for the purchase of a home. Through much of this US' history, however, blacks have been subject to a host of discriminatory practices in the housing market. In response to such discriminatory practices, Congress enacted the Fair Housing Act, Title VIII of the Civil Rights Act of 1968 (FHA). The unavailability of homeowners insurance in black neighborhoods contributes to the gap in homeownership rates. Through the FHA, Congress intended to make a broad statement against discrimination while leaving details to the future. Although some have argued that the insurance industry must be free to discriminate based on risk, this is no less true for the credit industry.

INTRODUCTION

President Bush has made homeownership a central element of his housing policy. Through the American Dream Down Payment Act, the Homeowner's Tax Credit, and the Zero Downpayment Mortgage Initiative, President Bush is attempting to make homeownership easier both by reducing the down payment required and by increasing financing options for the purchase of a home.3 More Americans owned homes in 2004 than ever before.4 This increase in homeownership has allowed more families to enjoy the "security, dignity, and independence that comes with owning their piece of the American dream."5

Homeownership is particularly important because of the numerous benefits it provides. It is one of the few vehicles through which low-income and minority families can accumulate capital." For this reason, it can provide a multigenerational path out of poverty. Not only do homeowners accumulate wealth that can be passed on to the next generation, homeownership increases psychological and physical well-being, self-esteem, and social capital." Children of homeowners are not only more likely to own homes themselves, but are also more likely to graduate from high school and college, are less likely to become pregnant while teenagers, are more likely to have greater future earnings, and are more likely to be involved in civic affairs.9

Through much of this nation's history, however, blacks have been subject to a host of discriminatory practices in the housing market. Realtors refused to show blacks homes in white neighborhoods. Lenders refused mortgages on the basis of race or the racial composition of neighborhoods. When blacks did succeed in buying homes, their investments often depreciated because realtors used blockbusting to increase housing prices temporarily in neighborhoods transitioning from white to black."'

In response to such discriminatory practices, Congress enacted the Fair Housing Act, Title VIII of the Civil Rights Act of 1968 (FHA)." Congress recognized that widespread racial discrimination in the housing market was preventing integration and interfering with minority access to jobs and quality education.12 Spurred by the race riots in the late 1960s, Congress looked to the FHA to ease the tension resulting from racial isolation.11 Through the FHA, Congress hoped "to eliminate the discriminatory business practices which might prevent a person economically able to do so from purchasing a house regardless of his race."14

Despite the many successes of the FHA, homeownership rates among blacks continue to lag behind those of whites; non-Hispanic whites are half again as likely as blacks to own their homes.15 The unavailability of homeowners insurance in black neighborhoods contributes to the gap in homeownership rates.16 Lenders refuse to write mortgages on uninsured property for fear of risking a destruction of their collateral simultaneous with a reduction of the borrower's resources and capacity for repayment. Insurers have been reluctant to write policies in black neighborhoods.17 When insurance is available, blacks pay more per dollar of insurance than do whites, even when controlling for income.18 The available data indicate that these premiums exceed those necessary to cover higher losses in black neighborhoods.

Insurers use a wide range of practices which contribute to the pricing differential in insurance for whites and blacks. For example, insurance companies typically establish a maximum age for houses beyond which policies become more expensive or entirely unavailable.20 Since a disproportionate number of blacks live in older homes,21 this results in making insurance unavailable or more expensive for this population.

Similarly, insurance companies consider the market value of a home in deciding whether to underwrite it and in deciding what type of policy to provide at what rate.22 For example, insurers often establish a minimumvalue threshold below which insurance is unavailable.23 Because low-income minorities need to buy inexpensive homes, and since insurance is necessary to secure a loan, minimum-value requirements operate to prevent poor people from buying homes. Insurers who do offer to cover inexpensive homes usually charge higher rates and refuse to write insurance for the entire replacement cost of the home.24

Insurance companies also adjust premiums by neighborhood in a way that increases prices in low-income, minority areas. Although insurance companies no longer draw red lines around minority neighborhoods to mark off areas where they will not underwrite policies,25 they accomplish similar results by cleaving territory based in part on crime rates, the percentage of owner-occupied homes in the neighborhood, the number of vacant buildings in the neighborhood, and the response time of the fire and police departments.26 Insurance companies fail to market insurance in low-income, minority areas27 and charge higher rates per dollar of insurance.28 As a result, insurance companies typically have lower loss-to-premium ratios in minority neighborhoods than in others.29

Finally, insurance companies use subjective criteria, such as requiring that the insured " 'be a person of integrity and financial stability who takes pride in his property.' "30 This language has been likened to "code words" used to exclude black customers in the era of overt redlining." Subjective criteria provide a vehicle for insurance agents to exercise their own biases.

This Note argues that because homeowners insurance is central to homeownership, the FHA applies to insurance underwriting policies, such as those mentioned above, that have a disparate impact on minority potential homeowners. Part I considers whether the FHA applies to homeowners insurance and concludes that homeowners insurance is covered by the Act. Part II goes on to argue that the FHA applies to homeowners insurance even where the discrimination results from disparate impact, rather than from disparate treatment. Finally, Part III analyzes the above-mentioned policies of the insurance industry under the FHA disparate impact standard.

I. APPLICATION OF THE FHA TO INSURANCE

Although courts have disagreed about whether the FHA applies to homeowners insurance, the great weight of authority supports its application.32 Section I.A analyzes the language of the FHA and concludes that it applies to insurance. Section I.B examines the legislative history and determines that applying the Act to insurance comports with congressional intent. Section I.C presents the alternative argument that, even if the FHA does not expressly apply to insurance, courts should give deference to the Department of Housing and Urban Development (HUD) regulation forbidding discrimination in insurance under the FHA. Finally, Section I.D determines that application of the FHA to insurance is consistent with state control of the insurance industry.

A. Statutory Interpretation

The language of the FHA should be interpreted to include homeowners insurance. Although insurance is not explicitly mentioned in the Act, the broad language of both § 3604 and § 3605 logically covers insurance.

Under § 3604(a), it is unlawful to do anything that makes a dwelling "unavailable."33 Because insurance is required in order to qualify for a mortgage, and since most people need a mortgage in order to buy a home, discrimination in underwriting decisions or in insurance pricing can make a dwelling unavailable, in contravention of § 3604.34

Although it could be argued that "otherwise make unavailable or deny"35 should only apply to activities similar to the refusal to sell or rent a home, not to all activities that make housing unavailable, this reading is inconsistent with other § 3604 jurisprudence. The Supreme Court has found that the FHA should be read broadly.36 Courts have readily applied § 3604 to a number of activities beyond the actual sale or rental transaction, such as zoning,37 the construction of low-income housing,38 and the provision of section 8 housing vouchers.39

Most courts have agreed that insurance, like zoning, is covered by § 3604. The Sixth Circuit explicitly rejected the argument that, under the canon of ejusdem generis, the FHA must be interpreted to exclude insurance.40 That court found a "direct connection of availability of property insurance and ability to purchase a house," which makes insurance discrimination the kind of discrimination banned by the FHA.41 The Seventh Circuit agreed that banning insurance discrimination forwards the FHA's goal of "removing obstacles to minorities' ownership of housing."42 Only the Fourth Circuit has disputed that this language applies to homeowners insurance.43 Various district courts have also held that this language applies to insurance, both in the context of racebased insurance redlining and in the context of underwriting hazard and liability insurance for mental health group homes.

Although most courts have analyzed insurance under § 3604, insurance also falls within the language of § 3605. Section 3605 makes it unlawful to discriminate in lending or in "providing other financial assistance."46 Since insurance is a prerequisite for obtaining a loan, the ban on discrimination in lending is meaningless if insurance discrimination is left unchecked. Thus, insurance is an integral part of the "residential real estate-related transactions" covered by § 3605 and logically falls within the scope of "financial assistance" mentioned in that section.47

The fact that insurance discrimination is a financial service within the scope of § 3605, however, does not mean that it is not also within the scope of § 3604.48 Statutory provisions commonly overlap,49 and, as explained above, insurance discrimination does function to "make unavailable or deny" housing within the scope of § 3604. Courts have analyzed other financial services arguably covered under § 3605, such as section 8 housing vouchers, under § 3604.50

B. Legislative History and Purpose of the FHA

The legislative history of the FHA demonstrates that applying the Act to insurance promotes Congress's policy goals. The purpose of the FHA, as expressed in the legislative history, is to prevent all discrimination relating to housing. The link between insurance and the purchase of a home puts insurance discrimination squarely within that purpose.

Through the FHA, Congress intended to make a broad statement against discrimination while "leaving details to the future."51 Congress made no attempt to specify all the services that would be covered by the Act. For example, the "Questions and Answers" document, used in the Senate, explains that the "Act will cover brokers, property owners, managers and anyone else who participates in the sale, rental or financing of housing." Similarly, as a result of negotiations in the Senate, the language of § 3605 in the final Act was broadened, making it unlawful not only to "deny a loan"53 but also to deny "other financial assistance."54 In introducing the Bill, one sponsor explained that it would "eliminate the discriminatory business practices" that affect homeownership.55 He did not mention any limits on the types of businesses affected.51

It is not surprising that insurance is not discussed in the FHA or its legislative history. The Fourth Circuit in Mackey v. Nationwide Insurance Co.57 reached the contrary conclusion that, because the insurance industry needs to be free to make decisions based on risk, Congress would have explicitly considered the needs of the insurance industry if it had intended the FHA to apply to insurance.58 The debates on the FHA, however, focused on the social problems created by housing discrimination, not on the policies through which such discrimination manifests.59 In introducing the Bill, its sponsors spoke extensively about the abysmal conditions in the "rotting cores of central cities,"60 such as inferior schools61 and the unavailability of jobs,62 but did not discuss specific housing practices that contributed to the problems.61 The only specific practice mentioned at all was blockbusting, which is expressly prohibited by the Act.64 Throughout the debates, the Senate focused on housing quality65 and overcrowding66 in inner-city, minority areas. Senators also argued that the Act could address concerns regarding the high price of housing in low-income, minority areas67 as well as racial differentials in homeownership rates.68 Insurance discrimination contributes to high costs and urban underinvestment,69 making it the type of discrimination that Congress was attempting to address through the FHA.

It could be argued that, in passing the FHA, Congress intended to remedy inner-city conditions only by making it possible for minorities to exit these areas, not by making financing available to those who remain in the inner city. Congress was very concerned with protecting minorities' rights to move to white neighborhoods.70 However, the supporters of the Act referred repeatedly to "choice" in housing.71 This must include the choice to invest in a home in the inner city, in addition to the choice to leave.72

That some members of Congress tried unsuccessfully to pass an amendment to the FHA to include insurance explicitly does not indicate that Congress did not intend the FHA to apply to insurance.7' The amendment was introduced by one of the original sponsors of the FHA, seeking to clarify the meaning of the bill in response to the Mackey decision.74 Its sponsor emphasized that the Act had always been intended to apply to all "links" in the "real estate chain," including homeowners insurance.75 The amendment was discussed only briefly in the Senate and never came up for a floor vote.76 The failure of legislation under these circumstances is likely unrelated to its merits.77 The Seventh Circuit considered this question and agreed that the failure of this amendment should not be considered determinative of Congressional intent.78

That Congress passed the Urban Property Insurance Protection and Reinsurance Act (UPIPRA) of 1968 the same year it enacted the FHA similarly does not imply that the FHA should not be interpreted to apply to insurance.79 Although the passage of insurance-related legislation the same year as the FHA might support an inference that Congress intended to handle insurance problems outside of the FHA, UPIPRA does not address the same issues as the FHA. The purpose of UPIPRA was to provide a subsidy to insurers in order to "protect private insurance companies from the risk of catastrophic losses which resulted from riots or civil disorders."80 UPIPRA did not address discrimination, the central goal of the FHA.81 Thus, the first court to consider the question determined that insurance redlining is the "especial province of the Fair Housing Act."82

C. Deference to HUD's Interpretation

Even if a court does not agree that the FHA clearly applies to insurance discrimination, courts should defer to HUD's interpretation, as expressed in 24 C.F.R. § 100.70(d)(4).81 Under Chevron U.S.A., Inc. v. National Resources Defense Council, Inc., courts defer to agency interpretations of their enabling statutes where (1) Congress has not "directly spoken to the precise question at issue" and (2) the "agency's answer is based on a permissible construction of the statute."85

Application of the FHA to insurance clearly survives the first prong of the Chevron test. According to the Chevron Court, the traditional tools of statutory construction should be used to determine whether Congress has directly spoken to the question.86 As explained in sections I.A-B, Congress has not directly spoken to the applicability of the FHA to homeowners insurance.87

Under the second prong of the Chevron test, courts should defer to HUD's reasonable interpretation of the statute as applying to insurance.88 HUD's interpretation is reasonable, as required under the second step of Chevron, because it is a possible interpretation of the language and forwards the purpose of the Act.89 The Sixth Circuit explicitly considered this question and concluded that HUD's interpretation is reasonable "in light of the direct connection of availability of property insurance and ability to purchase a house" and given the broad purpose of the FHA.90 The Seventh Circuit agreed that the application of the FHA to insurance forwards the goals of the Act and found that "(njothing in the text of the statute permits us to reject these proposed readings."91 The Fourth Circuit's ruling that the FHA does not apply to insurance was issued prior to the promulgation of 24 C.FR. § 100.70(d)(4) and prior to the Supreme Court's decision in Chevron.92

The consistency of HUD's position on the application of the FHA to insurance discrimination further indicates that deference is appropriate.93 On the basis of its expertise, HUD has consistently interpreted the FHA to cover insurance since the 1970's.94 HUD promulgated a regulation explicitly forbidding insurance discrimination under the FHA in 1989, less than a year after receiving rulemaking authority.95

HUD's expertise lends further credibility to its consistent determination that insurance discrimination can make housing unavailable. HUD has significant expertise in the issue of housing discrimination.96 Courts have recognized HUD's expertise when deferring to the agency's interpretations and decisions in other contexts.97 These cases give HUD considerable discretion to implement the varied and often competing goals of the national housing policy.98 Congress has also indicated its trust in HUD's expertise by delegating responsibility for interpretation of the FHA to HUD through its use of such broad language in the Act99 and by amending the FHA to give HUD rulemaking authority and increased enforcement powers.100 Given the consistency of HUD's position, the highly complex regulatory scheme of the FHA, the broad powers the Act confers on the agency, and the tremendous complexity of issues of residential segregation,101 HUD's determination that the FHA applies to insurance should carry great weight.102

D. Impact of the McCarran-Ferguson Act on the interpretation of the FHA

Insurance companies turn to the McCarran-Ferguson Act to argue that the FHA cannot be interpreted to apply to insurance.103 The McCarranFerguson Act provides that federal law does not preempt state insurance law unless the federal law "specifically relates" to insurance.104 Federal law not specifically relating to insurance should not be interpreted to "invalidate, impair, or supersede" state insurance law.'ni Some argue that, under McCarran-Ferguson, the FHA cannot be applied to insurance because it does not explicitly mention insurance and would preempt States' determinations of appropriate insurance discrimination regulations.106

The McCarran-Ferguson Act was enacted to handle conflicts between state insurance regulations and federal antitrust laws,107 and it should not be interpreted as invalidating subsequent civil rights laws. The Act was passed in 1945, in response to the Supreme Court's determination that insurance could be regulated under the Commerce Clause.108 Congress wanted to "preserve the traditional role of the states in the taxation and regulation of the business of insurance" and to exempt insurance companies from antitrust laws.109 Congress did not intend to exempt insurance companies from civil rights legislation.110 The Act's reference to the "business" of insurance and, specifically, to "fee[s] or tax[es]"111 demonstrates the Congressional focus on financial, not social, regulation of insurance companies.

Even if the McCarran-Ferguson Act is applied to civil rights legislation, federal law does not "invalidate, impair, or supersede" state law unless it explicitly conflicts with an express provision of state law.112 Every circuit court that has considered this issue has held that federal anti-discrimination laws do not conflict with state insurance laws.113 These cases typically require that the insurance company "point to" a particular law that would be invalidated, impaired, or superseded.114 The existence of a "general regulatory scheme" does not preclude the application of a federal antidiscrimination law, even where the federal law provides additional remedies.115 That the application of the FHA could affect insurance prices and interfere with states' control over prices does not mean that the state law is invalidated.116 Therefore, unless a particular state law allows insurers to discriminate contrary to the FHA, the McCarran-Ferguson Act does not apply.

II. APPLICATION OF THE FHA DISPARATE IMPACT STANDARD TO INSURANCE

In addition to banning disparate treatment in the insurance business, the FHA should also be interpreted to forbid insurance policies with a disparate impact that are not justified by business necessity. Section II.A defines the disparate impact test used in FHA litigation not relating to insurance. section II.B considers the applicability of that test to insurance. Section II.C argues that, as part of the showing of business necessity, the insurer should bear the burden of showing the lack of a less discriminatory alternative.

A. Disparate Impact under the FHA

While no appellate court has considered the application of the FHA to disparate impact insurance claims, the FHA does apply to disparate impact claims in contexts other than insurance. Eleven circuit courts have considered this question, and all agreed that disparate impact claims are viable under the FHA.117 Although the D.C. Circuit has not ruled on this point, it has indicated its agreement with the other circuits,118 and D.C. District Court judges have held that the FHA applies to disparate impact.119 The Supreme Court has construed similar language in Title VII to apply to disparate impact.120 The Court also remanded Arlington Heights v. Metropolitan Housing Development Corp.121 for consideration of FHA claims after finding that there was no discriminatory intent.122

Although there is some confusion regarding the appropriate test to apply to disparate impact cases under the FHA,123 a test derived from Title VII jurisprudence is appropriate in the FHA context. Under this test, in order to show a prima facie case of disparate impact, the plaintiff must (1) show that a disparate impact exists, (2) identify a policy alleged to have caused the disparate impact, and (3) show that the policy caused the disparate impact.124 If the plaintiff succeeds in making a prima facie case, the burden shifts to the defendant to show that the policy is justified by a business necessity.125 Even if the practice is justified by business necessity, the defendant is still liable if a less discriminatory alternative exists.

There is substantial support for this type of test among the courts. Several circuits have used this precise test.126 Courts that have adopted the alternative formulation agree that the primary considerations are (1) the existence of a discriminatory effect that is (2) not justified by a legitimate business necessity.127 As such, all circuits that have analyzed disparate impact under the FHA agree with the basic precepts of the test.128

The position of the Seventh Circuit further indicates the similarity of the tests. The Seventh Circuit introduced a different disparate impact test in Arlington Heights,129 one of the first disparate impact cases under the FHA.130 The circuit has since adopted the test described above, explaining that "[t]he analysis [in Arlington Heights] was in fact though not in words the 'disparate impact' analysis familiar from Title VII cases."131 The other courts applying the Arlington Heights test have not had the opportunity to revisit their rulings in light of the Seventh Circuit's subsequent analysis.132

B. Application of the FHA Disparate Impact Standard to Insurance

The language and history of the FHA support the application of its disparate impact test to claims of discrimination in insurance. The "because of race" clause in § 3604 applies equally to the sale or rental of a home and to other actions, such as insurance policies, that "make unavailable or deny" a home.133 Similarly, § 3605 makes discrimination "because of race" illegal in "transactions" and defines that term to include lending and "providing other financial assistance."134 Courts have held that the disparate impact standard applies to the other activities covered under § 3604 and § 3605.135 Given the integrality of insurance and housing136 and the broad goals of the FHA,137 it makes little sense to carve out an exception to the disparate impact standard for insurance.

Judicial decisions support the conclusion that the FHA applies to insurance policies with disparate impacts. Although no circuit court has considered a disparate impact challenge to insurance policy under the FHA,138 the broad language used to describe the applicability of the disparate impact standard does not suggest a distinction between insurance and other practices.139 When considering other suits under the FHA, courts have agreed that "[e]ffect, and not motivation, is the touchstone [under the FHA] . . . because . . . 'the arbitrary quality of thoughtlessness can be as disastrous and unfair to private rights and the public interest as the perversity of a willful scheme.'"140 Moreover, the three district courts that have considered disparate impact challenges to insurance practices have seen no reason to exempt insurance from the disparate impact test.141

Insurance companies maintain that the process of adverse selection in the insurance industry raises unique concerns, making a disparate impact test inappropriate for insurance.142 Insurance is priced based on the average risk of loss in the pool of insureds.143 Insurers argue that insurance pricing is most effective when insureds can be pooled into groups whose members share "similar characteristics."144 If risk levels were not similar among members of a pool, those whose risk levels were less than average for the pool would drop out rather than subsidize higher-risk members. This would increase the price of insurance for the remaining insured and decrease the profit for the insurance company. If left unchecked, this adverse selection would result in a "death spiral," in which only the highest-risk individuals would be willing to buy insurance, and the insurance market would "implode."145 Insurers argue that minority neighborhoods, particularly those in the inner city, are especially risky to insure because of high crime rates, poor services, and aging dwellings.146 If these factors were not taken into account, people in lower-risk neighborhoods would have to pay increased premiums to subsidize the costs of insuring high-risk, minority neighborhoods. This would result in raising prices beyond expected losses in low-risk neighborhoods, causing everyone who does not live in a high-risk minority neighborhood to drop their insurance coverage.147

The Fourth Circuit agreed with the insurance companies that, due to these special policy concerns, insurance cannot be lumped in with "other services" within the meaning of the FHA.148 The court found that, due to the special problem of adverse selection in insurance, the FHA does not apply even where insurers expressly consider race in underwriting decisions.149 Because of the special considerations in the insurance business, the court argued, Congress would not have made discrimination in homeowners insurance unlawful without special discussion.150 The Seventh Circuit, while applying the FHA to insurance disparate treatment, also suggested that the nature of the insurance industry makes disparate impact analysis inappropriate.151

Empirical evidence undermines this argument. Low-risk individuals do not reduce their coverage when placed in insurance pools with higher-risk insureds.152 This is because, first, people are very poor at calculating risk.153 The particularly high information and search costs relating to insurance exacerbate the impact of this weakness on insurance decisions.154 As a result, low-risk individuals, who could logically remove themselves from the insurance pool, will not recognize that they are paying more than their optimum rate.155 Second, the risk-averse are typically both more likely to buy insurance and less expensive to insure than others.156 This is because, in addition to purchasing insurance, these "belt-and-suspenders types" will take further steps to reduce their own risk. Since customers who are inexpensive to insure will not abandon the insurance market, insurers may not need special treatment under the FHA.

Moreover, even if insurance companies do need to be able to differentiate among risk categories to some extent, they need not be exempt from disparate impact liability. If the insurance industry actually requires the use of a particular indicator in order to operate profitably, it will prevail under the business necessity test.158 In a recent case, one district court applied a disparate impact test to insurance, dismissing the argument that insurance companies should not be held liable for disparate impact as "unavailing in light of the availability of the 'business justification' defense."159 In a similar case decided under Ohio state law, a court found that "the disparate-impact approach does not unduly undermine the business of selling insurance. Assuming . . . that the insurance industry is based on 'fair' risk discrimination, the disparateimpact approach will not impede such fair discrimination."160

The fact that circuit courts have held other financial services providers liable for disparate impacts further supports this conclusion.161 Lenders are similar to insurers in that both are concerned about the risk associated with their investment and look for accurate indicators of that risk. Although some have argued that the insurance industry must be free to discriminate based on risk,162 this is no less true for the credit industry. Therefore, since lenders are liable for disparate impacts, there is no reason to exempt insurers.163

C. Burden of Proof in the Disparate Impact Test for Insurance under the FHA

Courts disagree about who bears the burden of showing the absence of a discriminatory alternative, the final step of the disparate impact test. Some borrow from Title VII and hold that once the defendant has shown a business necessity, the burden should shift back to the plaintiff to rebut the defendant's evidence by showing that a less discriminatory alternative exists.164 Others require that the defendant show the lack of a less discriminatory alternative as part of the showing of business necessity.165 These courts argue that the Title VII test is inappropriate in the Title VIII context.166 Title VII's burden shifting is justified only by the need for employers to pick the best person for job, a concern not relevant when selecting a tenant.167

It is appropriate to place the burden on the insurer-defendant in Title VIII insurance discrimination cases.168 Not only does the Title VII justification not apply, but placing the burden on the defendant is justified in this context because only the insurer has access to data on the details of, and justifications for, its pricing models.169 Moreover, insurance companies routinely destroy data on rejected claims, making it particularly onerous for plaintiffs to prove discrimination.170 Since the defendant has sole access to the evidence, the defendant should also bear the burden of proof.171

Notwithstanding the protection afforded insurers by the business necessity test, insurance companies argue that placing the burden of proof on the insurance company effectively makes the business necessity defense unavailable. Insurance companies claim that requiring them to maintain information justifying the business necessity of each factor used in underwriting would be prohibitively expensive and damaging to the entire insurance industry.172 Although the data may not be easy for insurers to gather, they have much more ready access to this information than do their customers.173 If accurate classification of risk is as important to the insurance business as insurers claim, it seems unlikely that insurers would rely on risk-assessment factors without knowing how these factors actually correlate to risk. If, in fact, insurers are not maintaining this information, the business necessity requirement may aid them in improving their pricing models.174

III. APPLICATION OF THE DISPARATE IMPACT TEST TO SPECIFIC INSURANCE POLICIES

Many insurance pricing mechanisms disproportionately impact minorities. This Part applies the disparate impact test to insurers' use of age and value requirements, territory rating, and subjective factors such as good housekeeping and pride of ownership. Section III.A concludes that use of these policies establishes a prima facie disparate impact case. Section III.B considers the applicability of the business necessity defense in this context.

A. Elements of a Prima Facie Case of Disparate Impact

Establishing the first two elements of a prima facie case (the existence of specific policies and racial differences) is straightforward. Insurance companies have specific policies on the above-mentioned factors. A recent National Association of Insurance Commissioners (NAIC) report demonstrates the existence of a disparate impact with its finding that low-income, urban, minority homeowners are charged more for insurance and have fewer options than others.175 Therefore, this section concentrates on the third element: causation.

Proving that the maximum-age and minimum-value requirements cause a disparate impact should not be difficult, given the correlation between race and class. Researchers have found that minorities are more likely than whites to live in old homes.176 This is not surprising, as a higher percentage of blacks than whites live in poverty.177 It is common sense that poor people are more likely than others to buy inexpensive homes. In most areas, these inexpensive homes are likely to be old.178 Furthermore, since white-flight remains common,179 even middle-class blacks tend to live in less expensive homes than whites with equivalent incomes.180 Nationally, homes belonging to black homeowners are worth 18% less, on average, than homes owned by white homeowners with the same incomes.181 Some insurance companies have already eliminated maximum-age and minimum-value requirements, in recognition of their vulnerability to suit.182

Territorial rating practices should also be found to have a disparate impact.183 The specific factors considered by insurance companies, such as the number of owner-occupied homes in an area, the number of vacant buildings in an area, the crime rate, and the quality of municipal services, all vary between inner city and suburban areas.184 These factors have a disparate impact because residential segregation is still common,185 and minority homeowners are more likely than whites to own homes in inner-city areas.186

Perhaps because insurers recognize the vulnerability of these factors to disparate impact challenges, they have begun to rate territories based solely on the amount that the insurer has paid out in claims from that area.187 To establish that this type of territorial rating has a disparate impact, it is necessary to access the territorial ratings used by insurance companies. Unfortunately, limited data are available on this point because insurance pricing models have been protected as trade secrets.188 If minority areas receive more expensive territorial ratings than white areas, such ratings would have a disparate impact.

The use of subjective factors in determining risk may support racial discrimination because it allows insurance agents to act out their own biases. Although no comprehensive research has been done, there have been indications that agent bias has some effect. For instance, one study of homeowners insurance payments, following a hurricane in Florida, suggested that language and cultural barriers make assessors less likely to trust Latino insureds.189 Insurers are also more likely to require comprehensive inspection of homes in non-white neighborhoods.190 Agent bias has also been indicated by insurers' reluctance to market insurance in urban areas, even where they stand to profit. For example, in Missouri, the average commission per agent is $25,891 in areas with the highest minority population, compared to $3,081 in the areas with smallest minority population.191 Insurer bias could explain the unwillingness of agents to move into urban areas until this differential is equalized.192

Since 1996, a few lower courts have recognized that the above-mentioned policies cause sufficient disparate impact to support a prima facie case under the FHA. The D.C. District Court denied the defendant's summary judgment motion in a case challenging the use of (1) age, market value, and the difference between market value and replacement cost, (2) territorial rating zones that "reflect" racial composition, and (3) the use of credit ratings.193 In a similar case, an Ohio state court found sufficient evidence that using these factors is discriminatory to deny the insurer's motion for summary judgment. Both cases subsequently settled, with the insurance company agreeing to discontinue use of these factors and to increase marketing in urban areas.195 These outcomes provide further evidence of the causal link between these insurance practices and the disparate impact on minorities.

B. The Business Necessity Defense

The above-mentioned insurance policies are legal, in spite of their disparate impact, if justified by business necessity.196 The burden rests on the insurer to show that the metrics used to measure risk are predictive of loss. The insurer should also bear the burden of showing that the metrics used are better predictors of loss than other available metrics.197 For example, an insurer using the age of the home in underwriting and pricing decisions would have to provide data showing both that the age of a home is an accurate predictor of risk and that more specific indicators, such as the age of the wiring and plumbing, would not be more accurate.

The limited data available suggest that the use of factors that result in higher insurance rates for high minority areas may exceed what may be justified by business necessity.198 An analysis of data available in Missouri199 shows that, although homeowners in minority areas pay higher insurance rates than others,200 the ratio of losses paid by the insurance company to premiums paid by the insured is actually lower in high-minority areas (.73 compared to .81). This indicates that residents of minority areas are overcharged for their insurance.201 The one national study on point, a 1994 study of insurance data by rating territory done by NAIC, suggests that these differences are not unique to Missouri. The study concluded that "residents of low-income and minority neighborhoods face greater difficulty in obtaining high-quality homeowners insurance coverage" and pay more for that coverage.202 The NAIC report determined that the insurance industry had not provided loss-cost data to justify these differences.203

In all metropolitan areas for which data are available, the average insurance premium is substantially higher in neighborhoods where more than half of the residents are minorities than it is in neighborhoods where fewer than three percent of the residents are minorities.204 This is true even when controlling for neighborhood income level.205 In high-income white areas, premiums average 53% of those paid in high-income, high-minority areas.206 Similarly, residents of low-income white areas pay only 54% of the premiums paid by residents of low-income, minority areas.207 The difference is slightly less stark in medium-income areas, where residents of white neighborhoods pay 72% as much as residents in high-minority areas.208 The claim that these differences in pricing are needed to cover higher costs has not been substantiated by the Missouri data and the NAIC report.

CONCLUSION

When insurance policies function to make insurance more expensive in urban areas with large minority populations, the impacts are felt not only by the individual homeowners, but by broader society as well. Lack of insurance provides further justification for white flight, because lack of insurance in minority areas will make houses unmarketable.209 It hastens the deterioration of slums because landlords cannot invest in their properties, nor can purchasers buy empty buildings, without insurance. As The President's National Advisory Panel explained 38 years ago:

Insurance is essential to revitalize our cities. It is a cornerstone of credit. Without insurance, banks and other financial institutions will not-and cannot-make loans. New housing cannot be constructed, and existing housing cannot be repaired. New businesses cannot be opened, and existing businesses expand, or even survive.

Without insurance, buildings are left to deteriorate; services, goods, and jobs diminish. Efforts to rebuild our nation's inner cities cannot move forward. Communities without insurance are communities without hope.210

[Footnote]

1. Press Release, Office of the Press Secretary, Increasing Affordable Housing and Expanding Homeownership (Sept. 2, 2004), available at http://www.whitehouse.gov/news/releases/2004/ 09/20040902-5.html [hereinafter Expanding Homeownership].

2. 42 U.S.C.A. § 12821 (West 2005).

3. During the first six months of 2004, the Department of Housing and Urban Development (HUD) devoted $161.5 million to helping low-income, first-time homebuyers with down payments. U.S. DEP'T OF HOUS. & URBAN DEV., FISCAL YEAR 2006 BUDGET SUMMARY 7 (Feb. 7, 2005), http://www.hud.gov/about/budget/fy06/fy06budget.pdf.

4. U.S. Census Bureau, Housing Vacancy Survey-Historical Table 14, hltp:// www.census.gov/hhes/www/housing/hvs/historic/histt14.html (last visited May 28, 2006); U.S. Census Bureau, Housing Vacancies and Homeownership-Annual 2004: table 20, http:// www.census.gov/hhes/www/housing/hvs/annual04/ann04t20.html (last visited May 9, 2006).

5. Expanding Homeownership, supra note 1.

6. Julie Kosterlilz, Home Sweet Home?, 36 NAT'L J. 712 (Mar. 6, 2004) (" '[P]aying off the mortgage on a home has been, and will continue to be, the easiest way for (low-income and minority] households to build personal wealth.'"(quoting Stephen Brobeck)); Bruce Kalz et al., Rethinking Local Affordable Housing Strategies: Lessons from 70 Years of Policy and Practice (The Brookings Inst. & The Urban Inst., Discussion Paper, Dec. 2003) [hereinafter Brookings. Housing Policy] ("Housing wealth accounts for a vast majority of total assets for low-income and minority homeowners."). For example, for homeowners with household incomes below $20,000, homeownership accounts for seventy-two percent of net household wealth. For homeowners with incomes between $20,000 and $49,000. home equity constitutes fifty-five percent of total wealth. NICOLAS P. RETSINAS & ERIC S. BELSKY, LOW-INCOME OWNERSHIP: EXAMINING THE UNEXAMINED GOAL 201 (2002). Although some raise concerns that low-income households are more likely to be harmed by depreciations in home equity, e.g., Dean Baker, The Housing Bubble: A Time Bomb in Low-Income Communities?, NAT'L HOUSING INST., May-June 2004. http://www.nhi.org/online/issues/135/ bubble.html, the limited research available shows that low-price homes are more likely than expensive homes to increase in value and less likely to decrease in value. RETSINAS & BELSKY. supra, at 203-05.

7. Kosterlitz. supra note 6, at 714 ("[S]ome people see tackling racial disparities in homeownership as the next front in the battle for racial equality. 'The racial wealth gap is the starkest lever of racial inequality and is transmitted from one generation to the next.' " (quoting Thomas M. Shapiro)).

8. Brookings, Housing Policy, supra note 6, at 58-59. Although some of the studies attributing these benefits to homeownership are open to criticism on grounds that they may blur correlation and causation, low-income homeowners themselves attribute improvements in their lives to homeownership. William M. Rohe & Michael A. Stegman, The Effects of Homeownership on the Self-Esteem, Perceived Control and Life Satisfaction of Low-Income People, 60 J. AM. PLAN. ASS'N 173, 180 (2004) (finding that eight-five percent of low-income homeowners said that "owning a home made them feel better" and that seventy-one percent said that homeownership gave them "increased control over their lives").

9. Brookings, Housing Policy, supra note 6, at 59; see also Kosterlitz. supra note 6. Although homeownership in poor neighborhoods may produce fewer benefits, recent research shows that homeownership in almost any neighborhood will benefit children. Brookings, Housing Policy, supra note 6, at 60. In fact, children of renters may be better off if their parents purchase homes in distressed neighborhoods than if they remain renters and move to better neighborhoods. Id.

10. Blockbusting occurs when realtors sell homes to black families in a white neighborhood in such a way as to create a panic among the white residents. GEORGE R. METCALF, FAIR HOUSING COMES OF AGE 88 (1988). While residents leave the neighborhood and blacks, who are willing to pay extra to live in an integrated neighborhood, purchase the homes for a premium. Id. The resulting turnover in homes benefits realtors by increasing prices and sales. Id.

11. 114 CONG. REC. 2276, 2281 (1968).

12. Id. at 2273.

13. Charles L. Nier, III, Perpetuation of Segregation: Toward a New Historical and Legal Interpretation of Redlining Under the Fair Housing Act, 32 J. MARSHALL L. REV. 617, 628-30 (1999).

14. Nationwide Mut. Ins. Co v. Cisneros, 52 F.3d 1351. 1359 (6th Cir. 1995) (quoting Dunn v. Midwestern Indem. Mid-Am. Fire & Cas. Co.. 472 F. Supp. 1106, 1109 (S.D. Ohio 1979)).

15. Oversight of the Department of Housing and Urban Development, Including the Department's Budget Request for fiscal Year 2005: Hearing Before the H. Comm. on Financial Serv's., 108th Cong. (2004) (statement of Alphonse Jackson. Secretary, United States Department of Housing and Urban Development).

16. See PRESIDENT'S NATIONAL ADVISORY PANEL ON INSURANCE IN RIOT-AFFECTED AREAS, MEETING THE INSURANCE CRISIS OF OUR CITIES (1968) [hereinafter ADVISORY PANEL]; see also Cisneros, 52 F.3d al 1360 ("[T]he availability of property insurance has a direct and immediate affect on a person's ability to obtain housing.").

17. NAACP v. Am. Family Mut. Ins. Co., 978 F.2d 287, 297 (7th Cir. 1992).

18. See infra notes 199-203 and accompanying text.

19. Id.

20. Robert R. Detlefsen, Risk-Based Homeowners Insurance under Siege: The Slippery Slope from Redlining Charges to Disparate Impact Claims 11 (Nov. 1997), available at http://www.cei.org/pdf/1509.pdf.

21. Gregory D. Squires, Race, Politics, and the Law: Recurring Themes in the Insurance Redlining Debate, in INSURANCE REDLINING: DISINVESTMENT, REINVESTMENT, AND THE EVOLVING ROLE OF FINANCIAL INSTITUTIONS 1, 11 (Gregory D. Squires ed., 1997).

22. This information comes from informal conversations with agents at State Farm, American Family, Allstate, and Farmers Insurance companies.

23. Detlefsen, supra note 20, at 13.

24. Id.

25. This practice is known as "redlining." See, e.g., NAACP v. Am. Family Mut. Ins. Co., 978 F.2d 287 (7th Cir. 1992).

26. Detlefsen, supra note 20, at 11-14.

27. See, e.g., L. H. Otis, Availability Woes Documented in Pa., 101 NAT'L UNDERWRITER: PROP. & CASUALTY/RISK & BENEFITS MGMT., Sept. 15, 1997, at 3.

28. L. H. Otis, New Evidence of Urban Insurance Gap Released. 99 NAT'L UNDERWRITER: PROP. & CASUALTY/RISK & BENEFITS MGMT., June 26, 1995, at 3 [hereinafter Otis, Urban Gap].

29. See infra notes 199-203 and accompanying text. Recently, some insurance companies have replaced consideration of these factors with consideration of past loss rates from the neighborhood. Supra note 22. To the extent that the ahove factors are predictive of risk, using loss rates by neighborhood will continue to result in insurance costing more in poor, urban areas, hut may minimize differences in loss ratios.

30. Squires, supra note 21, at 6 (quoting a 1993 Nationwide Mutual Insurance Company underwriting manual).

31. Id.

32. Compare Nationwide Mut. Ins. Co. v. Cisneros, 52 F.3d 1351 (6th Cir. 1995), NAACP v. Am. Family Mut. Ins. Co., 978 F.2d 287 (7th Cir. 1992), Nat'l Fair Hous. Alliance, Inc. v. Prudential Ins. Co. of Am., 208 F. Supp. 2d 46 (D.D.C. 2002), Toledo Fair Hous. Ctr. v. Nationwide Mut. Ins. Co., 94 Ohio Misc. 2d 151 (Ohio C.P. 1997). and Dunn v. Midwestern Indem. Mid-Am. Fire & Cas. Co., 472 F. Supp. 1106 (S.D. Ohio 1979), with Mackey v. Nationwide Ins. Co., 724 F.2d 419 (4th Cir. 1984).

33. Section 3604 states, in relevant part:

it shall be unlawful

(a) To refuse to sell or rent after the making of a bona fide offer, or to refuse to negotiate for the sale or rental of, or otherwise make unavailable or deny, a dwelling to any person because of race, color, religion, sex, familial status, or national origin.

(b) To discriminate against any person in the terms, conditions, or privileges of sale or rental of a dwelling, or in the provision of services or facilities in connection therewith, because of race, color, religion, sex, familial status, or national origin.

42 U.S.C. § 3604 (2000) (emphasis added).

34. As Judge Easterbrook put it, "no insurance, no loan; no loan, no house." Am. Family, 978 F.2d at 297.

35. 42 U.S.C. § 3604 (2000).

36. Trafficante v. Metro. Life Ins. Co., 409 U.S. 205, 209 (1972); see also Michigan Prot. & Advocacy Serv., Inc. v. Babin, 18 F.3d 337, 344 (6th Cir. 1994) ("Congress intended § 3604 to reach a broad range of activities that have the effect of denying housing opportunities to a member of a protected class.").

37. Metro. Hous. Dev. Corp. v. Vill. of Arlington Heights, 558 F.2d 1283, 1290 (7th Cir. 1977).

38. Smith v. Town of Clarkton, 682 F.2d 1055 (4th Cir. 1982).

39. Langlois v. Abington Hous. Auth., 207 F.3d 43, 49 (1st Cir. 2000).

40. Nationwide Mut. Ins. Co. v. Cisneros. 52 F.3d 1351, 1357-58 (6th Cir. 1995).

41. Id. at 1359 (deferring to HUD's interpretation of the FHA as applying to insurance as promulgated in 24 C.F.R. § 100.70(d)(4) (2005)) (citing Dunn v. Midwestern Indem. Mid-Am. Fire & Cas. Co., 472 F. Supp. 1106, 1109 (S.D. Ohio 1979)).

42. NAACP v. Am. Family Mut. Ins. Co., 978 F.2d 287, 298 (7th Cir. 1992).

43. Mackey v. Nationwide Ins. Co., 724 F.2d 419, 423-24 (4th Cir. 1984).

44. Dunn, 472 F. Supp. at 1109.

45. See Wai v. Allstate Ins. Co., 75 F. Supp. 2d I (D.D.C. 1999) (applying the FHA to property insurance for group homes for the mentally retarded); Nevels v. W. World Ins. Co., 359 F. Supp. 2d 1110, 1117-22 (W.D. Wash. 2004) (extending the reasoning in earlier cases to apply the FHA to an insurer's refusal to provide liability insurance for group homes lor the mentally disabled).

46. Section 3605 states, in relevant part:

(a) In general

It shall be unlawful for any person or other entity whose business includes engaging in residential real estate-related transactions to discriminate against any person in making available such a transaction, or in the terms or conditions of such a transaction, because of race, color, religion, sex, handicap, familial status, or national origin.

(b) "Residential real estate-related transaction" defined

As used in this section, the term "residential real estate-related transaction" means any of the following:

(1) The making or purchasing of loans or providing other financial assistance

(A) for purchasing, constructing, improving, repairing, or maintaining a dwelling ...

42 U.S.C. § 3605 (2000) (emphasis added).

47. Id. This conclusion was reached by the court in National Fair Housing Alliance. Inc. v. Prudential Insurance Company of America. 208 F. Supp. 2d 46, 58 (D.D.C. 2002).

48. Insurers have argued that because S 3605 covers specific financial services, the phrases "otherwise make unavailable" and "discriminate ... in the provision of services" in § 3604 cannot be interpreted to refer to financial services. Mackey, 724 F.2d at 423. Because insurance is a financial service, and not explicitly mentioned in § 3605, it must be beyond the Act entirely. Id. The Fourth Circuit adopted this view in Mackey. Id.

49. NAACP v. Am. Family Mut. Ins. Co., 978 F.2d 287, 298 (7th Cir. 1992).

50. Langlois v. Abington Hous. Auth., 207 F.3d 43, 49 (1st.Cir. 2000).

51. Am. family, 978 F.2d at 299.

52. 114 CONG. REC. 2272 (1968) (emphasis added).

53. Id. at 2270.

54. Id. at 4572.

55. Id. at 2275.

56. Id.

57. 724 F.2d 419 (4th Gr. 1984).

58. Id. at 423.

59. NAACP v. Am. Family Mut. Ins. Co., 978 F.2d 287, 299 (7th Cir. 1992).

60. 114 CONG. REC. 2274 (1968).

61. Id. at 2273.

62. See, e.g.,id. at 2282.

63. See id. at 2270-81.

64. 42 U.S.C. § 3604(e) (2000).

65. See, e.g., 114 CONG. REC. 2526 (1968) (speaking of the goal of "a decent home . . . for every American family"); id. at 2528 ("[H]ousing of nonwhite families is consistently of poorer quality than that of white households of the same income level ... In I960, [forty-four] percent of all nonwhites lived in substandard housing as compared to [thirteen] percent of white families.").

66. See, e.g., id. at 2530, 2281.

67. Id. at 2528 ("[N]onwhites-whatever their income-pay higher prices for lower quality housing than white families."); id. at 2540 ("[T]here is a growing number of Negro Americans who are able to buy good housing, but who find that so much of their purchasing power is absorbed by exorbitant real estate costs in the ghetto . . . .").

68. Id. at 2529 (noting that rates of homeownership remain lower for nonwhites than for whites at every income level).

69. ADVISORY PANEL, supra note 16.

70. See. e.g., 114 CONG. REC. 2274 (1968) ("[H]ousing discrimination . . . tell[s] the Negro citizen trapped in an urban slum there is no escape, that even were he able to get a decent education and a good job, he would still not have the freedom other Americans enjoy to choose where he and his family will live.").

71. See. e.g., id. at 2283, 2524, 2998; Fair Housing Act of 1967: Hearing before the S. Comm. on Commerce and Finance, 90th Cong. 6 (1967) (statement of Ramsey Clark, Att'y Gen. of the United States).

72. This was recognized by the Second Circuit in banning racial quotas designed to maintain integrated communities by preventing white flight. United States v. Starrett City Assocs., 840 F.2d 1096, 1100 (2d Cir. 1988) ("Section 3604 'is designed to ensure that no one is denied the right to live where they choose for discriminatory reasons.'" (quoting Southend Neighborhood Improvement Ass'n v. County of St. Clair, 743 F.2d 1207, 1210 (7thCir. 1984))).

73. NAACP v. Am. Family Mut. Ins. Co., 978 F.2d 287, 299 (7th Cir. 1992).

74. 132 CONG. REC. 848 (1986).

75. Id.

76. See id.

77. See. e.g., Order of R.R. Conductors v. Swan, 329 U.S. 520, 528 (1947). This is particularly true where the tailed legislation reflects an attempt to clarify an agency's jurisdiction. See. e.g., FDA v. Brown & Williamson Tobacco Corp.. 529 U.S. 120, 182-83 (Breyer, J., dissenting); United States v. Sw. Cable Co., 392 U.S. 157, 170 (1968).

78. Am. Family, 978 F.2d at 300.

79. Nationwide Mut. Ins. Co. v. Cisneros, 52 F.3d 1351, 1358 (6th Cir. 1995).

80. Id. at 1358 (quoting Dunn v. Midwestern Indem. Mid-Am. Fire & Cas. Co., 472 F. Supp. 1106, 1111 (S.D.Ohio 1979)).

81. Id.

82. Dunn, 472 F. Supp. at 1112.

83. The HUD regulation makes it illegal to "[r]efus[e] to provide . . . property or hazard insurance for dwellings or providing such . . . insurance differently because of race, color, religion, sex, handicap, familial status, or national origin." 24 C.F.R. § 100.70(d)(4) (2005).

84. 467 U.S. 837 (1984).

85. Id. at 842-43. Although later decisions have somewhat limited the reach of the Chevron doctrine, they have not affected the application of the Chevron doctrine to rules, such as the HUD regulation at issue here, that were promulgated through notice and comment rulemaking procedures. See United States v. Mead Corp., 533 U.S. 218, 221 (2001 ) (denying Chevron deference to informal agency adjudication but preserving deference for notice and comment rules); Christensen v. Harris County, 529 U.S. 576, 587 (2000) (denying Chevron deference to interpretive rules); Bowen v. Georgetown Univ. Hosp., 488 U.S. 204 (1988) (denying Chemin deference to agency interpretations in legal briefs).

86. Chevron, 467 U.S. at 843 n.9.

87. If anything, Congress has indicated that the FHA does ban discrimination in homeowners insurance. See supra Sections I.A-B.

88. NAACP v. Am. Family Mut. Ins. Co., 978 F.2d 287, 298 (7th Cir. 1992).

89. See Barnart v. Thomas, 540 U.S. 20, 26 (2003) (relying primarily on the text of the Act to determine if the agency interpretation was reasonable); Babbitt v. Sweet Home Chapter of Cmtys. for a Great Or., 515 U.S. 687, 704-05 (1995) (relying on the legislative history and purpose of the statute, in addition to the text, to uphold the agency interpretation under of the Act); see also supra Sections I.A-B (arguing that the language and purpose of the Act indicate that it should be read to apply to insurance).

90. Nationwide Mut. Ins. Co. v. Cisneros, 52 F.3d 1351, 1359 (citing Dunn v. Midwestern Indem. Mid-Am. Fire & Cas. Co., 472 F. Supp. 1106. 1109 (S.D. Ohio 1979)).

91. Am. Family, 978 F.2d at 298.

92. See generally Mackey v. Nationwide Ins. Cos., 724 F.2d 419 (4th Cir. 1984).

93. See United States v. Mead Corp., 533 U.S. 218, 235 (considering consistency of the agency position as one factor indicative of the amount of weight that should be given to an agency interpretation).

94. Am. Family, 978 F.2d at 299-300.

95. This regulation makes it illegal to "[r]efus[e] to provide . . . property or hazard insurance for dwellings or providing such . . . insurance differently because of race, color, religion, sex, handicap, familial status, or national origin." 24 C.F.R. § 100.70(d)(4) (2005). The regulation was promulgated on January 23, 1989. Implementation of the Fair Housing Amendments Act of 1988, 54 Fed. Reg. 3232, 3232 (Jan. 23, 1989).

96. HUD is a well-established agency that has existed since 1965 and administers dozens of housing-related programs. HUD's History, http://www.hud.gov/library/bookshelf18/hudhistory.cfm (last visited Mar. 30, 2005). For a list of programs, see HUD Programs, http://www.hud.gov/ funds/index.cfm (last visited Mar. 30, 2005). For a discussion of the relevance of these factors in weighing the Agency's interpretation, see Mead, 533 U.S. at 234 (noting that agencies deserve additional deference where " 'specialized experience and broader investigations and information' [are] available to the agency" (quoting Skidmore v. Swift & Co., 323 U.S. 134, 139 (1944))); Aluminum Co. of Am. v. Cent. Lincoln Peoples' Util. Dist., 467 U.S. 380, 387 (1984).

97. Alschuler v. Dep't of Hous. & Urban Dev., 686 F.2d 472, 481-87 (7th Cir. 1982) (deferring to HUD informal adjudication regarding the location of housing under the section 8 program); Capitol Mortgage Bankers, Inc. v. Cuomo, 222 F.3d 151 (4th Cir. 2000) (deferring to HUD's interpretation of its authority in the area of mortgage lending).

98. See, e.g., Alschuler, 686 F.2d at 481-487.

99. Mead, 533 U.S. at 227 ("When Congress has 'explicitly left a gap for an agency to fill, there is an express delegation of authority to the agency to elucidate a specific provision of the statute by regulation,' and any ensuing regulation is binding in the courts unless procedurally defective, arbitrary or capricious in substance, or manifestly contrary to the statute." (quoting Chevron U.S.A. Inc. v. Nat'l Res. Def. Council, Inc., 467 U.S. 837, 843-44 (1984))). Congress intended to delegate this type of broad authority to HUD. One Senator criticized the FHA during the floor debate because it "empower|ed| the Secretary of HUD to do virtually everything." 114 CONG. REC. 2538 (1968).

100. Fair Housing Amendments Act, 42 U.S.C. §§ 3610-3612, 3614 (2005); Mead, 533 U.S. at 226-27 (holding that an administrative interpretation qualifies for Chevron deference when "Congress delegated authority to the agency generally to make rules carrying the force of law," which can be demonstrated by a grant of the "power to engage in adjudication or notice-and-comment rule-making," and the agency's interpretation is an exercise of that power).

101. See, e.g., STEPHEN GRANT MEYER, As LONG As THEY DON'T MOVE NEXT DOOR: SEGREGATION AND RACIAL CONFLICT IN AMERICAN NEIGHBORHOODS 11 (2000).

102. See Mead, 533 U.S. at 235 (considering these factors as indicative of the amount of weight that should be given to an agency interpretation).

103. This argument has been raised in every major insurance discrimination case. See, e.g., Dehoyos v. Allstate Corp., 345 F.3d 290, 293 (5th Cir. 2003); NAACP v. Am. Family Mut. Ins. Co., 978 F.2d 287, 293-94 (7th Cir. 1992).

104. The McCarran-Ferguson Act provides:

(a) State regulation. The business of insurance, and every person engaged therein, shall be subject to the laws of the several States which relate to the regulation or taxation of such business.

(b) Federal regulation. No Act of Congress shall be construed to invalidate, impair, or supersede any law enacted by any State for the purpose of regulating the business of insurance, or which imposes a fee or tax upon such business, unless such Act specifically relates to the business of insurance . . . .

15 U.S.C.A. § 1012 (West 2005).

105. Id. § 1012(b).

106. Almost every state has laws or regulations limiting discrimination in insurance. However, most of these regulations are much narrower than the FHA. Rosemary Baptiste et al., Redlining, Property Insurance and Urban Markets: Concepts, Issues, Initiatives, and Solutions, 49 SOC'Y CHARTERED PROP. & CASUALTY UNDERWRITERS 82, 96-102 (1996). State laws typically prohibit "unfair discrimination between individuals (if the same class or between neighborhoods within a municipality and of essentially the same hazard." See, e.g., COLO. REV. STAT. § 10-3-1104(1)(f)(II) (2004). Since insurance companies argue that wealthier, white neighborhoods do not represent the same level of hazard as poor, minority neighborhoods, these laws do not reach most discrimination in homeowners insurance.

107. Am. Family, 978 F.2d at 294.

108. United States v. Se. Underwriters Ass'n, 322 U.S. 533 (1944).

109. Mackey v. Nationwide Ins. Co., 724 F.2d 419, 420 (4th Cir. 1984).

110. Spirt v. Teachers Ins. & Annuity Ass'n, 691 F.2d 1054, 1065 (2d Cir. 1982), vacated tm other grounds, 463 U.S. 1223 (1983).

111. 15 U.S.C.A. § 1012 (West 2005).

112. See. e.g., Dehoyos v. Allstate Corp., 345 F.3d 290, 295-97 (5th Cir. 2003).

113. Id. at 299; Moore v. Liberty Nat'l Life Ins. Co., 267 F.3d 1209, 1221-22 (11th Cir. 2001) (considering preemption of § 1981 and § 1982); Nationwide Mut. Ins. Co. v. Cisneros, 52 F.3d 1351, 1363 (6th Cir. 1995); Merchants Home Delivery Serv., Inc. v. Frank B. Hall & Co., 50 F.3d 1486, 1491-92 (9th Cir. 1995) (considering preemption of RICO); NAACP v. Am. Family Mut. Ins. Co., 978 F.2d 287, 297 (7th Cir. 1992); Mackey; 724 F.2d at 421.

114. Dehoyos, 345 F.3d at 297.

115. Mackey, 724 F.2d at 421.

116. Dehoyos, 345 F.3d at 297 n.5. But see McClain v. Shelter Gen. Ins. Co., No. 97-1139-CV-W-FJG. 2005 U.S. Dist. LEXIS 34662, at *19-20 (W.D. Mo. Mar. 31, 2005) (holding that the court should not issue an order that would alter insurance rales approved by the state commission, in pan because the commissioner considered discrimination in approving rates).

117. See, e.g., Simms v. First Gibraltar Bank, 83 F.3d 1546, 1555 (5th Cir. 1996); Mountain Side Mobile Estates P'ship v. Sec'y of Hous. & Urban Dev., 56 F.3d 1243. 1250-51 (10th Cir. 1995); Jackson v. Okaloosa County, 21 F.3d 1531, 1543 (11th Cir. 1994); Casa Marie, Inc. v. Superior Court of Puerto Rico, 988 F.2d 252, 269 n.20 (1st Cir. 1993); United States v. Starrett City Assocs., 840 F.2d 1096, 1100 (2d Cir. 1988); Keith v. Volpe, 858 F.2d 467, 482-84 (9th Cir. 1988); Arthur v. City of Toledo, 782 F.2d 565, 575 (6th Cir. 1986); Smith v. Town of Clarkton, 682 F.2d 1055, 1065 (4th Cir. 1982); United States v. Mitchell. 580 F.2d 789, 791-92 (5th Cir. 1978); Resident Advisory Bd. v. Rizzo, 564 F.2d 126. 147-48 (3d Cir. 1977); Metro. Hous. Dev. Corp. v. Vill. of Arlington Heights, 558 F.2d 1283, 1290 (7th Cir. 1977); United States v. City of Black Jack, 508 F.2d 1179, 1184 (8th Cir. 1974).

118. Miller v. Poretsky, 595 F.2d 780, 790 n.10 (D.C. Cir. 1978).

119. See, e.g., Nat'l Fair Hous. Alliance, Inc. v. Prudential Ins. Co. of Am., 208 F. Supp. 2d 46, 58-59 (D.D.C. 2002).

120. Griggs v. Duke Power Co., 401 U.S. 424, 432 (1971) (construing the phrase "because of race," which also appears in the FHA).

121. 429 U.S. 252, 271 (1977).

122. The Supreme Court has continued to leave open the question of disparate impact under the FHA after a number of courts ruled in favor of a disparate impact test. City of Cuyahoga Falls, Ohio v. Buckeye Cmty. Hope Found., 538 U.S. 188, 199-200 (2003) (ordering that disparate impact claims should he dismissed because plaintiffs "abandoned" them without commenting on the applicability of disparate impact theories to the FHA); Town of Huntinglon v. Huntington Branch, NAACP, 488 U.S. 15, 18 (1988) ("Since appellants conceded the applicability of the disparate-impact test for evaluating the zoning ordinance under Title VIlI, we do not reach the question whether that test is the appropriate one. Without endorsing the precise analysis of the Court of Appeals, we are satisfied on this record that disparate impact was shown . . . .").

123. See, e.g., Peter E. Mahoney, The End(s) of Disparate Impact: Doctrinal Reconstruction, Fair Housing and Lending law, and the Antidiscrimination Principle, 47 EMUKV L.J. 409, 411 (1998) (describing the standard as "sketchy and haphazard").

124. See, e.g., Resident Advisory Bd. v. Rizzo, 564 F.2d 126, 148-49 (3d Cir. 1977).

125. 42 U.S.C. §2(k)(1)(A) (2000). Several FHA cases were decided between the Supreme Court's decision in Wards Cove Packing Co. v. Atonio, 490 U.S. 642 (1989). and the reinstatement of the "business necessity" test by Congress in the Civil Rights Act of 1991. These cases apply the more lenient "business justification" test of Wards Cove. However, these cases should not he given weight, as courts deciding FHA claims alter 1991 have agreed that the "business necessity" test applies. John F. Stanton, The Fair Housing Act and Insurance: An Update and tin· Question of Disability Discrimination, 31 HOFSTRA L. REV. 141, 186-87 (2002).

126. See Fair Hous. in Huntington Comm. v. Town of Huntington, 316 F.3d 357, 366 (2d Cir. 2003); Langlois v. Abington Hous. Auth., 207 F.3d 43. 49-50 (1st Cir. 2000); Harris v. Itzhaki, 183 F.3d 1043, 1051 (9th Cir. 1999) (explaining that "[w]e apply Title VII discrimination analysis in examining Fair Housing Act discrimination claims," although the decision was ultimately based on a finding of discriminatory intent) (quoting Gamble v. City of Escondido, 104 F.3d 300, 304 (9th Cir. 1997))); Mountain Side Mobile Estates P'ship v. Sec'y of Hous. & Urban Dev., 56 F.3d 1243, 1250-51 (10th Cir. 1995); Rizzo, 564 F.2d at 148-49; United States v. City of Black Jack, 508 F.2d 1179, 1184-85 (8th Cir. 1974); Owens v. Nationwide Mut. Ins. Co., No. 3:03-CV-1184-H. 2005 U.S. Dist. LEXIS 15701, at *49-50 (N.D. Tex. Aug. 2, 2005). The Eighth Circuit did not mention the policy and causation elements, requiring only that "the conduct of the defendant actually or predictably results in racial discrimination." However, that case considered a zoning ordinance, so the policy and causation elements were clear. City of Black Jack, 508 F.2d at 1184.

127. Stanton, supra note 125, at 186.

128. Other circuits, although recognizing disparate impact under the FHA, have not had occasion to formulate precise tests, see Simms v. First Gibraltar Bank, 83 F.3d 1546, 1555 (5th Cir. 1996); Jackson v. Okaloosa County, 21 F.3d 1531, 1543 (11th Cir. 1994); Hanson v. Veterans Admin., 800 F.2d 1381. 1389 (5th Cir. 1986) (finding inadequate proof of causation and not discussing the burden-shifting portion of the test), although a lower court in the Fifth Circuit has adopted the Title VII test. Owens, 2005 U.S. Dist. LEXIS 15701, at *49-50.

129. Metro. Hous. Dev. Corp. v. Vill. of Arlington Heights, 558 F.2d 1283, 1290 (7th Cir. 1977).

130. The Seventh Circuit found four factors to be relevant in determining whether disparate impact was present:

(1) how strong is the plaintiff's showing of discriminatory effect; (2) is there some evidence of discriminatory intent, though not enough to satisfy the constitutional standard of Washington v. Davis; (3) what is the defendant's interest in taking the action complained of; and (4) does the plaintiff seek to compel the defendant to affirmatively provide housing for members of minority groups or merely to restrain the defendant from interfering with individual property owners who wish to provide such housing.

Id. Finding only the first and fourth factors present, the court called it a "close case" and decided in favor of the plaintiffs. Id. at 1290-93.

131. Vill. of Bellwood v. Dwivedi, 895 F.2d 1521. 1533 (7th Cir. 1990).

132. See Smith v. Town of Clarkton, 682 F.2d 1055, 1065 (4th Cir. 1982); Arthur v. City of Toledo, 782 F.2d 565, 575 (6th Cir. 1986) (adopting three of the four factors but declining to consider intent).

133. 42 U.S.C. § 3604 (2000).

134. Id. § 3605.

135. See, e.g., Simms v. First Gibraltar Bank, 83 F.3d 1546, 1555 (5th Cir. 1995) (loans); United Stales v. City of Blackjack, 508 F.2d 1179, 1184 (8th Cir. 1974) (rental).

136. See supra note 16 and accompanying text.

137. See supra notes 59-72 and accompanying text.

138. The Fifth and Sixth Circuits have both avoided ruling on disparate impact insurance discrimination cases. Dehoyos v. Allstate Corp., 345 F.3d 290 (5th Cir. 2003) (addressing only the McCarran-Ferguson preemption issue on interlocutory appeal prior to the settlement of the case); Nationwide Mut. Ins. Co. v. Cisneros, 52 F.3d 1351, 1363 (6th Cir. 1995) (finding that, in the absence of a specific HUD regulation on the applicability of the disparate impact test to insurance, the question was not ripe for review, but finding against the insurance companies on disparate treatment grounds).

139. See, e.g., Simms, 83 F.3d at 1555 (holding that a "violation of the FHA may be established . . . by a showing of significant discriminatory effect" without any language limiting the holding to certain types of violations).

140. City of Black Jack, 508 F.2d at 1185 (quoting Hobson v. Hansen, 269 F. Supp. 401, 497 (D.D.C. 1967)).

141. Owens v. Nationwide Mut. Ins. Co., No. 3:03-CV-1184-H. 2005 U.S. Dist. LEXIS 15701, at *49-50 (N.D. Tex. Aug. 2, 2005); Nat'l Fair Hous. Alliance, Inc. v. Prudential Ins. Co. of Am., 208 F. Supp. 2d 46. 49 (D.D.C. 2002); Toledo Fair Hous. Ctr. v. Nationwide Mut. Ins. Co., 94 Ohio Misc. 2d 151 (Ohio C.P. 1997).

142. Although insurers make similar arguments against the application of the FHA to discriminatory treatment, the Mackey court is the only court ever to find the argument compelling in the discriminatory treatment context. Mackey v. Nationwide Ins. Co., 724 F.2d 419 (4th Cir. 1984).

143. Detlefsen, supra note 20, at 4.

144. NAACP v. Am. Family Mut. Ins. Co., 978 F.2d 287, 290 (7th Cir. 1992).

145. Peter Siegelman. Adverse Selection in Insurance Markets: An Exaggerated Threat, 113 YALE L.J. 1223, 1254 (2004).

146. Detlefsen, supra note 20, at 9.

147. Id. at 5.

148. Mackey v. Nationwide Ins. Co., 724 F.2d 419, 423 (4th Cir. 1984).

149. Id.

150. Id. This argument should tail when applied to direct racial discrimination. The insurance industry does not require insurers to "engage in disparate treatment, to draw lines on the basis of race rather than risk." NAACP v. Am. Family Mut. Ins. Co., 978 F.2d 287, 298 (7th Cir. 1992). As Judge Easterbrook observed, "[r]isk discrimination is not race discrimination." Id. at 290. Prejudice would undermine, not improve, the accuracy of insurance pricing. To the extent that the success of the insurance business relies on accurate risk assessment, refusal to follow the FHA's ban on disparate treatment would harm the insurance industry.

151. Am. Family, 978 F.2d at 290.

152. Siegelman, supra note 145, at 1248.

153. Id. at 1241-46 (surveying studies that demonstrate human inability to calculate their own risk of loss).

154. Squires, supra note 21, at 76.

155. Id.

156. Siegelman, supra note 145, at 1270-74 (surveying studies that demonstrate propitious selection, rather than adverse selection, in insurance). For example, people who buy life insurance live longer than those who do not, even after controlling for factors such as income and smoking status. Id. at 1270.

157. Id. at 1266; see also Amy Finkelstein & Kathleen McGarry, Private Information and Its Effect on Market Equilibrium: New Evidence from Long-Term Care Insurance 3, 30 (Nat'l Bureau of Econ. Research, Working Paper No. 9957, 2003), available at http://papers.nber.org/papers/ w9957.pdf (finding that more cautious individuals are more likely to have long-term care insurance and less likely to enter a nursing home because they are more likely to seek preventative care).

158. Cf. Thomas v. First Fed. Sav. Bank, 653 F. Supp. 1330, 1341 (N.D. Ind. 1987) (approving use of a loan-to-value ratio in underwriting decisions as a legitimate business criterion, even though it makes loans unavailable on low-value, inner-city homes).

159. Nat'l Fair Hous. Alliance, Inc. v. Prudential Ins. Co. of Am., 208 F. Supp. 2d 46, 60 (D.D.C. 2002).

160. Toledo Fair Hous. Ctr. v. Nationwide Mut. Ins. Co., 94 Ohio Misc. 2d 151, 157 (Ohio CP. 1997).

161. See, e.g., Simms v. First Gibraltar Bank, 83 F.3d 1546, 1555 (5th Cir. 1995).

162. See, e.g., Prudential Ins., 208 F. Supp. 2d at 60.

163. This is particularly true in today's economy, since many companies provide both loans and insurance to their customers. Many insurers even have mortgage lenders as subsidiaries. William C. Apgar & Mark Duda, The Twenty-Fifth Anniversary of the Community Reinvestment Act: Past Accomplishments and Future Regulatory Challenges, 9 FED. RES. BANK N.Y. ECON. POL. R. 169, 183 (2003). However, the application of disparate impact to credit continues to be controversial because it is believed to interfere with the assessment of credit risk. See generally Mahoney, supra note 123.

164. See Oti Kaga, Inc. v. S.D. Hous. Dev. Audi., 342 F.3d 871, 883-84 (8th Cir. 2003); Mountain Side Mobile Estates P'ship v. Sec'yof Hous. & Urban Dev., 56 F.3d 1243, 1254 (10th Cir. 1995) (applying the Title VII standard without discussion of possible differences in the Title VIII context).

165. See Fair Hous. in Huntington Comm. v. Town of Huntington, 316 F.3d 357, 366 (2d Cir. 2003); Resident Advisory Bd. v. Rizzo, 564 F.2d 126, 148-49 (3d Cir. 1977).

166. See, e.g., Rizzo, 564 F.2d at 148-49 (differentiating Title VII cases and placing the burden of proving that no less discriminatory alternative exists on the defendant in FHA cases because the qualities that employers may legitimately consider are more easily defined than those in the housing context, and because "the consequences of an error in admitting a tenant do not seem nearly as severe as, for example, the consequences of an error in hiring an unqualified airline pilot" (quoting Elliot M. Mineberg, Comment, Applying the Title VII Prima Facie Case to Title VIII Litigation, 11 HARV. C.R.-C.L. L. REV. 128, 174 (1976))).

167. Id.

168. See, e.g., Stanton, supra note 125, at 184 (arguing that defendants are due less deference in Title VIII cases than in Title VII cases because there is less need for the defendant to make fine distinctions between applicants; employers may be harmed by not hiring the best person for the job, but these considerations do not apply in the housing context).

169. William E. Murray, Homeowners Insurance Redlining: the Inadequacy of Federal Remedies and the Future of the Property Insurance War, 4 CONN. INS. L.J. 735, 744-45 (1997-98).

170. Id.

171. This is comparable to the justification of the doctrine of rex ipsa loquitur. Cf. Byrne v. Boadle, 159 Eng. Rep. 299, 301 (Exch. Div. 1863) (placing the burden of proof on the defendant because "how could [the plaintiff] possibly ascertain from what cause it occurred?").

172. Detlefsen, supra note 20, at 38-39.

173. See supra notes 169-170 and accompanying text.

174. Cf. Squires, supra note 21, at 76 (arguing that "[r]esearch on the causes of urban property losses . . . could further help insurers to improve the efficiency and fairness of their pricing systems and underwriting criteria.").

175. Otis, Urban Gap, supra note 28, at 3.

176. Squires, supra note 21, at 11.

177. CARMEN DENAVAS-WALT ET AL., U.S. CENSUS BUREAU, PUBL'N No. P60-226, INCOME, POVERTY, AND HEALTH INSURANCE COVERAGE IN THE UNITED STATES: 2003, at 9 (2004), available at http://www.census.gov/prod/2004pubs/p60-226.pdf (finding that the poverty rate among blacks was 24.4 percent in 2003, compared with 8.2 percent among whites).

178. MEYER, supra note 101, at 8. As urban areas grow, upper- and middle-income residents typically move into new homes on the urban periphery. ANTHONY DOWNS, NEIGHBORHOODS AND URBAN DEVELOPMENT 38-39 (1981). The value of older homes left behind in the center is depressed by the availability of new homes on the periphery, and these homes are occupied by poor households. Id.

179. MEYER, supra note 101, at 218-19.

180. Brookings, Housing Policy, supra note 6, at 58.

181. Id. For every dollar of income, white owners had $2.64 worth of house. Black owners had $2.16 worth of house. Id.

182. L.H. Otis, Top Carriers Make Moves in Urban Market, 100 NAT'L UNDERWRITER: PROP. & CASUALTY/RISK & BENEFITS MGMT. ED. 1 (1996). The use of maximum age has been replaced with other indicators such as the condition of roof and the age of wiring, furnace, and plumbing. Barbara Bowers, Redeveloping the Urban Market. 99 BEST'S REV. 11, 32 (1999); see also Stanton, supra note 125, at 176 (discussing the use of lead paint as a proxy for age of home).

183. See supra notes 25-29 and accompanying text (discussing specific territorial rating policies).

184. For example, in 2004, 53.1% of people in central cities owned their homes, compared to 75.7% of suburban residents. U.S. Census Bureau, Housing Vacancies and Homeownership, http://www.census.gov/hhes/www/housing/hvs/annual04/ann04t12.html (last visited May 17, 2006). Crime rates in inner-city areas typically exceed those in suburban areas. In 2003, for example, an urban home was half again as likely as a suburban home to be burglarized, Bureau of Justice Statistics. Victimization Rates by Type of Crime and Locality of Residence, http:// www.ojp.usdoj.gov/bjs/pub/pdf/cvus/current/cv0353.pdf (last visited May 17, 2006), and black, urban heads of household were twice as likely as white, suburban heads of household to be victims of burglary. Bureau of Justice Statistics, Victimization Rates By Locality of Residence, Race of Head of Household and Type of Crime, http://www.ojp.usdoj.gov/bjs/pub/pdf/cvus/current/ cv0355.pdf (last visited May 17, 2006).

185. See MEYER, supra note 101, at 11.

186. Brookings, Housing Policy, supra note 6, at 56.

187. Supra note 22.

188. The California Supreme Court recently ordered State Farm to release the number of policies issued and canceled by zip code. In doing so. the court rejected the argument that these data should be protected as trade secrets, on grounds that the intent of Proposition 103 was to override trade secret claims. State Farm Mut. Auto. Ins. Co. v. Garamendi. 32 Cal. 4th 1029, 1046-47 (2004). This decision suggests the potential for providing access to additional information.

189. Tom Baker & Karen McElrath, Whose Safety Net?: Home Insurance and Inequity, 21 LAW & Soc. INQUIRY 229, 254 (1996). This study also found indications that insurance companies pay men more in damages than they pay women. Id. at 253.

190. Murray, supra note 169, at 737.

191. Squires, supra note 21, at 88.

192. Id. at 74.

193. Nat'l Fair Hous. Alliance Inc. v. Prudential Ins. Co. of Am., 208 F. Supp. 2d 46, 49 (D.D.C. 2002). The court accepted, for the purposes of the 12(b)(6) motion, the plaintiffs' allegation that these policies were "not justified or supported by business necessity . . . and [that] less restrictive, non-discriminatory alternatives [were] available to meet any legitimate business objectives." Id. at 50.

194. Toledo Fair Hous. Ctr. v. Nationwide Mut. Ins. Co., 94 Ohio Misc. 2d 151 (Ohio C.P. 1997) (considering a disparate impact insurance claim under provisions of the Ohio Civil Rights Act similar to HUD regulation 24 C.F.R. § 100.70(d)(4) (2005)).

195. Toledo Fair Hous. Ctr. v. Nationwide Mut. Ins. Co., 94 Ohio Misc. 2d 186, 195-203 (Ohio C.P. 1998).

196. See supra note 125 and accompanying text.

197. See supra Section II.C.

198. NAIC and the Missouri Department of Insurance have made public limited information which suggests insurance companies pay out the most per dwelling in neighborhoods where more than half of the residents are minorities. Interestingly, the data also show that loss costs are lowest in integrated neighborhoods (with minority populations of four to fifty percent). Squires, supra note 21, at 58.

199. Missouri is the only state that has maintained comprehensive data on losses and premiums by zip code over a long period. Id. at 53. Unfortunately, national data are not currently available.

200. Homeowners in high minority areas pay higher insurance rates than others, averaging $5.06 per thousand dollars of insurance, compared to $4.37 in areas where fewer than three percent of the residents are minorities. Id. at 59.

201. Id. The difference is even greater for limited coverage insurance plans, Id.

202. Otis, Urban Gap, supra note 28, at 3.

203. Id.

204. Squires, supra note 21, at 56.

205. Id.

206. Id.

207. Id.

208. Id.

209. The "[d]iscriminatoiy denial of access to 'normal' insurance and relegation of minorities to state FAIR plans creates 'the racially segregated housing patterns' which the Fair Housing Act is designed to prevent." Dunn v. Midwestern Indem. Mid-Am. Fire & Cas. Co., 472 F. Supp. 1106, 1111-12 (S.D. Ohio 1979).

210. ADVISORY PANEL, supra note 16, at 1.