The battle over the use of credit scores when underwritinginsurance risks is reheating to a fever pitch following the releaseof a controversial Federal Trade Commission report hailed by theindustry but hammered by consumer groups and a dissenter within theagency. Credit scores are an “effective predictor” of risk underauto insurance policies, the FTC concluded in its report, butcounterattacks have undermined the study's credibility.

Industry trade groups praised the study–”Credit-Based InsuranceScores: Impact on Consumers of Automobile Insurance”–with officialsof the National Association of Mutual Insurance Companies, forexample, echoing other property-casualty organizations in statingthat “the report confirms what the industry has been saying allalong–that credit-based insurance scores provide an objective andreliable tool” for setting insurance rates.

Opponents of the technique argue that credit scoring unfairlyimpacts low-income and minority groups, and fails to account forthose who deal in cash or are impacted by one-time events, such asmedical emergencies.

However, the report–mandated by the Fair and Accurate CreditTransactions Act (FACT Act) of 2003–concluded that the use ofcredit-based insurance scores actually “may result in benefits forconsumers.”

“For example, scores permit insurance companies to evaluate riskwith greater accuracy, which may make them more willing to offerinsurance to higher-risk consumers, for whom they would otherwisenot be able to determine an appropriate premium,” according to thereport.

“Scores also may make the process of granting and pricinginsurance quicker and cheaper–cost savings that may be passed on toconsumers in the form of lower premiums,” the report adds.

However, the report cautions, “little hard data was submitted oravailable to quantify the magnitude of these benefits toconsumers.”

At the same time, the report concedes that credit-basedinsurance scores are distributed differently among racial andethnic groups, and this difference is likely to have an effect onthe insurance premiums these groups pay, on average.

Specifically, the report says, non-Hispanic whites and Asiansare distributed relatively evenly over the range of scores.

However, African-Americans and non-white Hispanics aresubstantially overrepresented among consumers with the lowestscores–those associated with the highest predicted risk–”andsubstantially underrepresented among those with the highestscores.”

The agency also said even though it tried a variety ofalternative approaches, “the FTC was not able to develop analternative credit-based insurance scoring model that wouldcontinue to predict risk effectively yet decrease the differencesin scores on average among racial and ethnic groups.”

The agency said this doesn't mean a model could not beconstructed to meet both objectives, “but it does strongly suggest,however, that there is no readily available scoring model thatwould do so.”

To the relief of carriers, the report concluded that, ingeneral, credit scores “are predictive of the number of claimsconsumers file and the total cost of those claims. The use ofscores is therefore likely to make the price of insurance bettermatch the risk of loss posed by the consumer. Thus, on average,higher-risk consumers will pay higher premiums and lower-riskconsumers will pay lower premiums.”

Carl Parks, NAMIC's senior vice president of government affairs,said “the practice encourages competition, enables insurers tooffer coverage to more consumers at a fair price, and helpsstreamline the decision-making process.”

June Holmes, interim CEO for the Property Casualty InsurersAssociation of America, said the report “reaffirms the strongconnection between credit information and the risk of loss, and hasdetermined that its use helps to increase the availability andaffordability of insurance for most consumers.” She added that “nowthere should be no doubt about the value of using this highlypredictive underwriting and rating tool.”

David Snyder, vice president and assistant general counsel atthe American Insurance Association, said the study “confirms whatwe have long professed and many previous studies have shown–thatcredit-based insurance scores help refine insurance pricing tobetter reflect an individual's risk profile. We believe scoresreduce subsidization of bad risks by good ones, meaning mostconsumers pay less for insurance.”

Mr. Snyder added that while “the report is very detailed andrequires more thorough analysis…the FTC's initial conclusion saysit all–'credit-based insurance scores are effective predictors ofrisk under automobile policies.'”

Consumer groups were outraged by the report and disputed itscredibility. Indeed, a consortium of four such organizations calledon Congress to reject the FTC's work and order the agency toconduct “an objective, independent study.”

The groups went on to say that “based on the available evidenceof racial discrimination, Congress should ban the use of insurancecredit scoring.”

The FTC study is fatally flawed because the insurance industrycontrolled data used in the analysis, according to a jointstatement released by the Consumer Federation of America, theNational Fair Housing Alliance, the National Consumer Law Center,and the Center for Economic Justice.

“Instead of requiring the submission of comprehensive policydata by a large number of insurers, the FTC used data handpicked bythe insurance industry,” the groups charged in their statement.

“Buried in the report,” the organizations said, is the fact thatthe FTC noted in its study that “the alleged correlation betweenrisk and credit-based insurance scores might be explained by otherfactors.”

“To add insult to injury, the FTC report mimics the insuranceindustry's blaming-the-victim psychobabble of claiming credithistory is related to responsibility and risk management,” saidShanna L. Smith, National Fair Housing Alliance president and chiefexecutive officer.

“A look at the actual scoring models shows that socio-economicfactors have more impact on the score than loan payment history,and that an insurance credit score has little to do with personalresponsibility and everything to do with economic and racialstatus,” she added.

Indeed, the groups said “the relationship between insurancecredit scores and race is so strong that even though the FTC useddata handpicked by the industry, it found that credit scoringdiscriminates against low-income and minority consumers, and thatinsurance scoring was a proxy for race.”

J. Robert Hunter, insurance director of for the ConsumerFederation of America, said he was “disappointed” by the FTCreport, adding that its conclusions were “kind of expected” givenhow the study was conducted. “They didn't go after real data. Theyjust took what the industry gave them.”

Mr. Hunter added there were some positives to the report–mostimportantly that the FTC could not provide an explanation of whycredit scores affect a consumer's risk. “It confirms there's nothesis,” he said, adding that the study is “just correlating a dataset.”

By the same token, Mr. Hunter argued, insurers could use anynumber of other factors that have been shown to correlate withrisk–specifically mentioning a study conducted in California whichhe said showed that the darker a consumer's hair color, the greaterrisk they are for insurers. “Why shouldn't we use hair color [tounderwrite insurance]?” he asked.

Additionally, Mr. Hunter said that it was “important” that thereport found a disaproportionate effect on consumers based onrace.

Meanwhile, the report was also attacked by a dissenting memberof the FTC. Commissioner Pamela Jones Harbour disagreed with themethodology used in generating the report and voiced doubt over the“reliability of any conclusions the report might draw.”

In her dissent, Ms. Harbour charged that the “data collectionand analysis fell short of the FTC's gold standard for rigor andcompleteness, and did not reflect the agency's best practices.” Sheadded that “better alternatives were available and should have beenutilized.”

Ms. Harbour noted that “had this report been based on the realinsurance marketplace–using actual, verifiable data on individualpolicyholders, from a broad cross-section of insurancecompanies–reliable answers might have emerged.”

She said she couldn't endorse the report “due to my gravemethodological concerns. This study fell short of the rigorousresearch and data-collection standards to which the Commissionusually adheres.”

A second FTC member–Commissioner Jon Leibowitz–cited otherconcerns, saying that while he thought the report “makes asubstantial contribution to public discussions in this area,” theresults “are, of course, no cause for celebration.”

The internal criticism prompted FTC Chair Deborah Platt Majorasand the two other supporting commissioners to issue a statementdefending the report.

“While we respect the dissent's views as to the data andmethodology used here, we have confidence in the quality of theprocess that the Commission staff used and the soundness of theresults obtained,” the report's three supporters said.

A hearing to discuss the FTC's report scheduled for July 27before the House Subcommittee on Oversight and Investigations,chaired by Rep. Mel Watt, D-N.C., was unexpectedly postponed, withno new date set as this story went to press.

“The growing use of credit-based insurance scores affectsmillions of consumers,” Rep. Watt said. “We look forward to hearingfrom a range of perspectives–including state insurance regulators,consumer advocacy groups and the insurance industry–on thisimportant issue.”

The chief of staff for the Florida Department of Insurance saidone reason the hearing was postponed was because Florida InsuranceCommissioner Kevin McCarty informed the committee he had a personalemergency and had to return home as soon as possible. He wasscheduled to represent the National Association of InsuranceCommissioners at the hearing.