Root Insurance Vows to Stop Using Credit Scores to Set Rates

Written By
G. Dautovic
February 09,2021

Root Insurance recently announced it’s ambitious plan to eliminate the use of credit scoring as a factor in its car insurance pricing model.

The use of non-driving factors like credit scores in calculating auto insurance premiums has long been a subject of debate by industry leaders who believe in its necessity and consumers who find these practices unjust.

According to a survey conducted by Root, two thirds of Americans don’t even know that credit scores affect car insurance rates. Meanwhile, 93% said the industry should work to remove bias and discrimination from its pricing.

“Eliminating credit scores is a major and necessary step toward dismantling archaic industry practices and making car insurance fairer,” said Alex Timm, co-founder and CEO of Root. “We are committed to working with our partners, regulators and industry stakeholders to adopt this important change, and hope our announcement today inspires others to join us in fighting discrimination, bias and systemic inequity in auto insurance. It’s time to drop the score.”

Currently, 47 states allow the use of credit factors in auto insurance underwriting, while more than 90% of US insurers engage in the practice. Root said in a press release that this and other demographic factors have disproportionately affected historically under-resourced communities, immigrants, and those struggling to pay large medical expenses, so much so that drivers with safe records and low credit scores are penalized for $1,500 or more in annual premiums.

The company previously removed other similarly biased factors from insurance ratings, such as occupation and education, but eliminating the use of credit scores will be a far more daunting task.

Root said it needs at least five years to fully remove the scoring from its pricing method and increase reliance on telematics and other factors. The timetable is designed to help the company avoid a market disadvantage and ensure that the new methodology is as accurate as the current one.

"The reality is credit grouping is still correlated—not causal, nor accurate at the individual level—with insurance costs," a Root spokesperson explains. "Any information that's predictive of insurance costs will be used by the market due to the economic forces at play; anyone who doesn't use that data might be at a large disadvantage unless they can test and find more predictive information to use in place. Telematics will serve as that information, but we have to demonstrate the predictive and fair nature of that data."

Another big hurdle is convincing regulators to understand and accept the use of telematics, especially since each state has its own regulatory body.

“This isn’t easy,” said Tom Kuhn, director of communications at Root. “Credit scores have long been viewed by the industry and regulators as one of the most predictive indicators of risk. There is a lot of work to be done to implement the changes, get approval for those rates on a state-by-state basis and phase-in any impact to our policyholders.”

Kuhn added that driving behavior measured through telematics will be more predictive and fairer than credit scores when it comes to insurance rates. He called on other insurers to join the company’s fair car insurance stance.

Root hopes that other insurers will follow and answer a recent call from the National Association of Insurance Commissioners to address discrimination through research, discourse, and eventual recommendations to increase diversity and inclusion in the industry.
About author

I have always thought of myself as a writer, but I began my career as a data operator with a large fintech firm. This position proved invaluable for learning how banks and other financial institutions operate. Daily correspondence with banking experts gave me insight into the systems and policies that power the economy. When I got the chance to translate my experience into words, I gladly joined the smart, enthusiastic Fortunly team.

More from blog