Credit-based Insurance Scoring Remains a Key Legislative, Regulatory Issue

By | November 7, 2005

Credit-based insurance scoring has been a contentious legislative and regulatory issue in the world of insurance for at least the last three or four years, but the nature of the debate has changed, Robert Detlefsen, National Association of Mutual Insurance Companies, director of public policy, told the Society of Insurance Research at their annual conference, Oct. 19, in St. Petersburg, Fla.

“Insurers increasingly turned to consumer credit scores as a basis for underwriting and pricing auto and homeowners insurance coverage after numerous studies established a correlation between an individual’s credit score and the probability that he would file an auto or homeowners insurance claim,” Detlefsen said. “Initially, some policymakers and consumer activists were skeptical of the relationship between credit history and future loss because most of the original studies were sponsored by the insurance industry.

“Moreover, some critics argued that even if the correlation exists, insurers hadn’t adequately explained the reason for it,” Detlefsen continued. He said that whether a cause-and-effect relationship exists is immaterial from an underwriting standpoint.

“The only thing that matters is whether credit history accurately predicts future losses,” he said.

Detlefsen observed that in the face of recent government studies offering “incontrovertible evidence of credit scoring’s actuarial validity,” critics have attempted to discredit the practice by arguing that it discriminates against minorities and the poor. He noted that class-action discrimination lawsuits have been filed against insurers in Texas and Louisiana, and that some regulators and lawmakers have moved to ban credit-based insurance scoring on the grounds that it is racially discriminatory.

Insurers, however, do not collect information regarding the race or ethnicity of their customers or applicants for insurance coverage. “Since credit-based insurance scoring is colorblind and the insurers that use it are oblivious to race, where does the charge of racial discrimination come from?” he asked.

Adverse statistical impact
Detlefsen explained that the idea that credit scoring is unfairly discriminatory relies on the so-called disparate impact theory of discrimination, which originated in a 1971 U.S. Supreme Court case, Griggs vs. Duke Power.The Duke Power Company of North Carolina required job applicants to possess a high school diploma and pass a written test to be considered for employment. He said that although the policy applied to all job applicants regardless of race, it had the effect of excluding a higher percentage of black applicants than white applicants. In other words, even though many blacks met the requirements and were hired (and many whites didn’t meet the requirements and weren’t hired), blacks were statistically overrepresented among applicants who didn’t meet the company’s standards.

The Supreme Court responded to these facts by ruling that the “disparate impact” caused by the company’s employment criteria violated Title VII of the Civil Rights Act of 1964. Detlefsen said the decision astonished legal experts and employers at the time, because Title VII simply states that employers cannot discriminate “because of” one’s race or ethnicity. “It says nothing about disparate impact or effect,” Detlefsen said. “Essentially, the Civil Rights Act banned disparate treatment based on these characteristics. It says that you cannot treat people differently on the basis of race in making employment decisions.”

Application to insurance underwriting
While courts have largely confined use of the disparate impact doctrine to employment discrimination cases, some insurance industry critics have advanced the novel idea that it should be applied to credit-based insurance scoring, pointing to studies indicating that blacks and Hispanics have lower credit scores on average than whites and Asians.

Detlefsen said it would be “wrong and counterproductive” to apply disparate-impact analysis to credit scoring and other actuarially valid underwriting variables, in part because to do so would deprive consumers of all races with good credit scores of the benefits of lower rates.

He said the situation would be different if insurers subjected consumers to disparate treatment based on race and ethnicity.

Redlining?
According to Detlefsen, the disparate-impact argument against credit scoring comes in several guises. He noted that critics sometimes claim credit scores are merely a “proxy” for race. “If this were true,” he observed, “one would be able to identify an individual’s race or ethnicity simply by looking at his credit score. That’s what it means to say that one characteristic is a proxy for some other characteristic. Needless to say, it’s impossible to tell someone’s race based solely on his credit score.”

Detlefsen said that “redlining,” an emotionally-charged term in the world of insurance, has also been disingenuously invoked by critics of credit-based insurance scoring. “Credit scoring would constitute a very peculiar form of redlining,” he said. “While certain minority groups are statistically overrepresented among people with low credit scores, in terms of absolute numbers, the largest cohort of people with low credit scores is non-Hispanic whites. That’s not surprising since non-Hispanic whites make up roughly 70 percent of the population.”

Detlefsen explained that an insurer that used credit scoring as a redlining technique “would discover that most of the people who paid higher rates or were not offered coverage because of poor credit scores belong to the very group that the insurer is supposedly trying to favor. At the same time, the insurer would be charging lower rates to the countless minority consumers who have good credit scores. The redlining canard makes no sense in the context of credit scoring.”

From lawsuits to regulation
Detlefsen warned that if the disparate impact theory of discrimination were to migrate from the courts to the legislative and regulatory arena, “the insurer would lose the opportunity to demonstrate that the use of credit scores in underwriting and pricing serves a legitimate business purpose.”

Unlike a legal proceeding where defendants can present evidence to rebut an inference of unfair discrimination based on a disparate impact claim, a law or regulation that categorically prohibited credit-based insurance scoring would simply assume that the practice serves no legitimate purpose, he said.

“That’s contrary to everything we know about credit scoring’s ability to predict future losses,” Detlefsen concluded. “Once a law is enacted and on the books, the debate is over and it simply doesn’t matter that credit scoring is an actuarially valid underwriting tool.”

Was this article valuable?

Here are more articles you may enjoy.

From This Issue

Insurance Journal West November 7, 2005
November 7, 2005
Insurance Journal West Magazine

PLUS Conference Issue