Disparate Impact Dangers
On March 10 one of the nation's largest property and casualty insurers, Nationwide Insurance Company, agreed to part with $13.2 million to settle allegations of property insurance "redlining" brought by the U.S. Department of Justice. Less sensationally - but perhaps more significantly - Nationwide also agreed to substantially revise its standard underwriting criteria with respect to the urban homeowners market. No longer will Nationwide make underwriting decisions on the basis of such objective factors as the age or market value of a home.
The Nationwide "settlement" is the most recent instance of a major insurance company succumbing to government pressure to change its underwriting practices for urban property insurance. Allstate, State Farm, and American Family agreed to do so earlier, after the federal Department of Housing and Urban Development and "fair housing" activists accused the insurers of inner-city redlining. While the sudden flurry of federal lawsuits and settlements is a new development, the redlining issue itself is not. For nearly three decades, activists have campaigned to limit insurance companies' use of risk-based methods for underwriting and pricing homeowners insurance. But the movement has entered an ominous new phase in the 1990s.
Redlining once meant refusing to insure properties located in certain geographic regions. That is still the case insofar as state insurance regulation is concerned, where redlining has been universally prohibited for decades. In recent years, however, redlining has been elevated to the level of a federal civil rights violation through the expedient of "disparate impact," a specious legal doctrine, which holds that a standard or practice is presumptively illegal if it has the effect of disproportionately excluding members of certain minority groups - even though the challenged practice makes no reference to race or ethnicity, and even though the resulting adverse group impact was inadvertent.
By casting redlining as a form of racial discrimination, fair housing groups have succeeded in mobilizing the federal government's formidable civil rights enforcement apparatus to overturn traditional risk-based underwriting standards. In December 1993, then-HUD Secretary Roberta Achtenberg dispatched a memo to officials in the agency's Office of Fair Housing and Equal Opportunity advising that "[insurance redlining] cases which have been brought under the Fair Housing Act should now be analyzed using a disparate impact analysis. . ." The tactic led to HUD's claiming, inter alia, that insurers who refuse to offer replacement-cost policies for older homes are guilty of illegal discrimination, since older homes are more prevalent in minority neighborhoods.
Insurers, it must be conceded, do indeed "discriminate" among prospective customers on the basis of risk. So it is hardly surprising that compared to other homeowners, residents of predominantly minority inner-city neighborhoods often pay more for homeowners insurance, while frequently receiving less coverage. Compared to other properties, inner-city homes are more likely to incur losses because they tend to have substandard heating and wiring components (which is generally a function of age), and are located in neighborhoods with higher crime rates, more abandoned buildings, and a greater incidence of arson.
Despite the recent spate of redlining settlements, as yet there is not an actual decision that establishes a legal precedent for applying the disparate impact doctrine to insurance underwriting. But that could soon change. A pending federal court case, Canady v. Allstate Insurance Co., et al., would use disparate impact analysis to invalidate traditional risk-based underwriting standards employed by 23 individual insurance companies. An amicus brief filed by the Justice Department strongly endorses the use of disparate-impact analysis.
If HUD, DOJ, and their clientele of federally-funded fair housing advocates succeed in formally codifying the disparate-impact approach to allegations of insurance redlining, the result will be a radical transformation of the nature of homeowners insurance. One can imagine a future in which insurers will be required to document a precise cause-and-effect relationship between each underwriting and pricing variable they use and its associated risk. They will then be required to show that no "less discriminatory" risk-assessment technique is available.
Where it is not possible or too costly to meet this burden, insurers would have no choice but to abandon the use of those risk selection practices and cost-based pricing mechanisms that yield a disparate racial impact. Insurers would then be forced to distribute the higher, more frequent claim costs of one group of homeowners among another group of homeowners who present a lower level of risk, in effect creating a cross-subsidy. That would inevitably lead to what insurers call "adverse selection," the phenomenon whereby low-risk insureds tend to purchase less coverage, and high-risk insureds tend to purchase more coverage than they would if prices were more accurate.
How slippery is the disparate-impact slope upon which the insurance industry has been launched? Consider this: Disparate-impact enthusiasts are now taking aim at insurers who have set up shop on the Internet. According to Gregory Krohm, editor of the Journal of Insurance Regulation, "the profile of the average Web user is an insurer's dream come true. They're well-educated, relatively affluent people, young to middle-aged. And you can reach these people at very low cost."
But this could prove something of a two-edged sword in a world where anything that produces a disparate impact among groups is inherently suspect. A recent editorial in the National Underwriter urges regulators to "watch out for 'Weblining' in Internet marketing," and warns that "the same disparate impact bug" that bit insurers in the federal redlining cases will "buzz all around their Web sites if their cyberspace strategies aren't implemented for the benefit of all."
The editorial does not pause to consider that for an insurer to do this, it would need to provide free computers, training, and Internet access accounts to those in need. So the practical effect of applying disparate impact theory to Internet insurance marketing would be to prohibit the practice altogether. That would be too bad for consumers who, as the Philadelphia Inquirer reported recently, "find that going online is a convenient, efficient, low-stress way to determine insurance needs and price policies." These people will be denied the benefits of Internet marketing, simply because some other consumers lack the inclination or the means to access computer technology. And the disparate-impact juggernaut rolls on.