Consumer Group Claims NAMIC Study on Credit Scoring ‘Flip-Flops’ Earlier Positions

The consumer group, Center for Economic Justice (CEJ), released a statement in response to a National Association of Mutual Insurance Companies (NAMIC) study on credit scoring.

CEJ’s Executive Director Birny Birnbaum said, “The insurance industry cannot reverse itself fast enough on its credit scoring positions,” also stating that claims in the NAMIC study “are proven to be false.”

Below is CEJ’s response to the study, The Legal Theory of Disparate Impact Does Not Apply to the Regulation of Credit-Based Insurance Scoring, which offers dissecting arguments to statements in the NAMIC study.

The full NAMIC study can be downloaded from NAMIC’s Web site at: www.namic.org. (Also see today’s news for NAMIC’s response to CEJ’s critque.)

Relationship between Credit Scores, Race and Income

According to CEJ, in 1998 and 1999, the American Insurance Association presented a “study” to the National Association of Insurance Commissioners (NAIC) claiming that “credit scoring
is not significantly correlated with income.”

“The insurance industry also routinely cited a 1999 report from the Virginia Bureau of Insurance (VBI), claiming the report concluded ‘that neither race nor income were reliable predictors of scores’ – a claim the VBI denies,” said the statement.

The statement said: “That from 1998 through 2002, the insurance industry routinely presented these studies as proof that credit scoring did not have a disproportionate impact on poor and minority consumers.”

The statement continued. “For example, the testimony before the Georgia Insurance Commissioner in October 2001 and testimony before almost every state legislature during the period. At the same time the industry was denying any link between credit scorig and race or income, the industry was stonewalling insurance regulators and refusing to provide the data necessary for an independent analysis of the issue. ”

The CEJ statement claims that “as more information came out about the credit scoring models, organizations like CEJ started documenting the clear relationship between credit scoring factors and income and the large number of credit scoring factors related to economic status instead of payment history.”

CEJ statement added that, “contrary to insurer claims, the University of Texas study actually showed that credit scoring was a proxy for other factors already used by insurers and was correlated to race. A few states – Maryland, Washington and Alaska – tried to examine whether credit scoring was a proxy for prohibited factors, such
as race or income. But when the Missouri Department of Insurance issued its study earlier this year – a rigorous statistical analysis using state-of-the-art techniques – showing wide disparity in credit scores by income and race, the industry had to change course.”

“Today, the industry wants to characterize any criticism of credit scoring as “a disparate theory” that has no place in insurance. The industry mischaracterization of the multi-state credit scoring study is but one example,” the statement claimed.

Credit Scoring Study Standards

According to the CEJ statement, the industry has relied upon a simple “univariate” analysis of credit scoring and insurance losses to justify the use of credit scoring. Univariate means that insurers compared credit scoring to one factor – loss ratios

“Despite the fact that in 1996, the NAIC labeled such analyses ‘counterproductive and misleading,’ the industry continued to rely upon these studies through 2003. But, as soon as the Missouri Department of Insurance study was released, the industry criticized that study because it relied upon univariate analyses! (In fact, the Missouri study also employed multivariate analysis, but that inconvenient detail was omitted in the industry criticism.),” the CEJ statement said.

“But the flipping and flopping was just beginning.” the statement continued. “Now the industry had once again hired their favorite actuary – Mike Miller – to produce a study justifying industry practices. And now the industry proclaimed a new standard for actuarial studies – multivariate analysis that simultaneously considers risk classification factors other than credit scoring to “control” for the impact of those other factors and attempts to isolate the impact of credit scoring on losses. We will put aside the fact that the NAIC had called for this type of analysis since 1996, but the industry refused year after year to provide regulators with the necessary data. Apparently, the industry only felt comfortable having their in-house actuaries do the job. And we should also put aside the fact that industry hailed the University of Texas study – despite that being a univariate analysis.”

According to CEJ, “the industry fought tooth and nail to prevent the NAIC from performing a comprehensive study of credit scoring. The industry argued that any credit scoring study must include a review of losses and not just the relationship between credit scores and race or income. The industry criticized the Missouri study because it did not include loss data – only correlations between credit scores and race or income. The industry now rejected univariate analysis and demanded multivariate analysis as the correct approach.”

“So the states involved in the multi-state study issued a data call to consider losses and other rating factors in a multivariate analysis to determine if credit scoring is predictive of losses beyond the factors already used by insurers and to determine if credit scoring is a
proxy for prohibited factors such as race,” the statement said.

“As such, the multi-state study is firmly grounded in the rating laws – which prohibit unfair discrimination – of every state. Now the industry had a problem. The regulators were doing what the industry requested and moving to perform a detailed multivariate analysis – just like what the industry had asked Mike Miller to do. How could the industry derail this latest effort by regulators?” the statement asked.

“In the time-honored tradition of misinformation and mischaracterization – by claiming that the multi-state study was something different that it was and then criticizing their distortion. Now the industry characterized the multi-state study as a “disparate impact” study and then mounted a campaign against “disparate impact. Hence, the NAMIC “policy” paper,” said the statement.

According to the CEJ statement, there are “false statements, unsupported allegations and myths” in the NAMIC study.

“The NAMIC paper continues the insurance industry tradition of false statements and unsupported allegations regarding credit scoring and related issues,” the CEJ statement said.

“Disparate impact is a legal standard that has not been applied to insurance,” as stated in the NAMIC study (see page 7, executive summary).

According to CEJ statement, “This statement would be surprising to National Fair Housing Alliance and its member organizations who filed fair housing complaints against insurers for unfair discrimination
based on the disparate impact legal standard – a position that was upheld in numerous court cases. There is no longer any serious legal debate – the NAMIC fantasy papernotwithstanding – that federal civil rights laws apply to residential property insurance. The latest ruling along these lines was in the case of DeHoyos vs. Allstate, where the court allowed the lawsuit challenging Allstate’s use of credit scoring as unfair discrimination to go forward.”

“The majority of consumers benefits from credit scoring and pays less,” as stated in the NAMIC study (see page 8, executive summary).

According to the CEJ statement, “When coupled with the fact that credit scoring penalizes consumers in low-income and minority communities, this is an interesting argument from insurers. If we assume that the claim is true, then what insurers are saying is that unfair discrimination is okay as long as the majority benefit. By that logic, why not charge African Americans higher rates than other races? Since African Americans are a minority of the population, the majority of consumers would benefit. This argument is profoundly un-American. This industry allegation is completely unsupported. Insurers have routinely denied regulators – at least the few regulators interested enough to ask – the data necessary to test this claim. Michigan is an exception and when the Michigan Office of Financial and Insurance Services examined this claim, they found that the majority of consumers would benefit from a ban on credit scoring – and Gov. Granholm took action to effect that ban.”

“Insurance scoring allows companies to write more business,” as stated in the NAMIC study (see page 8, executive summary).

According to the CEJ statement, “Again, the insurers make this claim without an iota of evidence to support it. In fact, hundreds of agents attest to the fact that credit scoring prevents them from writing
business in working class and minority communities – the same business they used to write before insurers started using credit scoring. If credit scoring allowed insurers to write more business, then why are so many agents groups opposed to credit scoring, including the National State Farm Agents Association, the National Association of Professional Allstate Agents and the United Farmers Agents Association? Simply stated, their agents groups would not oppose credit scoring if it allowed them to write more
business.”

“Application of the disparate impact theory to insurance underwriting erodes the moral consensus on which the nation agreed to abolish racial discrimination in the Civil Rights Act of 1964,” stated in the NAMIC study (see page 8, executive summary).

According to the CEJ statement, “Here the industry is saying that different insurance outcomes for different races – higher prices and lower availability for poor and minority consumers – are okay as long insurers use a proxy for race – instead of race itself – to charge minority consumers more. This is the same industry that used age and value of properties to deny coverage in older, economically disadvantaged communities for decades. The insurers argued that they had data to show that older and lower-valued homes had higher losses than other properties. Fair housing organizations argued that insurers were engaging in unfair discrimination and redlining entire older inner-city neighborhoods. When the insurers finally settled the
litigation against fair housing organizations – which used the disparate impact legal standard of federal civil rights laws – the insurers agreed to stop using these guidelines and admitted they would now write more business in economically-disadvantaged communities. Under the insurer theory of “civil rights,” cited above from the NAMIC
paper, the fair housing groups would not have been able to lodge their complaint and consumers in poor and minority communities would continue to be redlined.”

Birnbaum said, “It is only in the world of insurance executives where raising someone’s auto or homeowners insurance rates because he or she has been laid off from a job is considered fair. In the real world, raising someone’s insurance rates because they’ve been the victim of a medical or economic catastrophe is unfair.”

“At a time when state insurance regulators are scrambling to defend state insurance regulation and fend off a federal government takeover, it is extremely disappointing to see only a handful of state insurance regulators willing to stand up to the industry they regulate. How many times can the insurance industry stonewall and deceive regulators
and policymakers before the insurance commissioners start testing industry claims and developing independent information for legislators and the public?”