Many insurance buyers understand that credit score can affect their ability to secure a car loan or to qualify for a mortgage. However, most consumers do not realize just how much credit score affects the price they pay for insurance.
The use of credit score in pricing insurance goes back to the mid-1990s, when insurance companies realized this was a new ‘silver bullet’ that could be used to more accurately price insurance risk. Credit score first gained traction for use in auto insurance, where the hypothesis was that drivers with higher credit scores are more likely to maintain their vehicles and avoid claims. It was also argued that those who are more ‘responsible’ in their financial affairs are more likely to exhibit ‘responsible’ driving behavior.
rates paid by two homeowners with identical attributes other than credit score can be 80% to 200% (3 times) higher
Within 10 years, by the middle of the 2000s, virtually all auto insurers were using credit score to price auto insurance. Insurance carriers slotted customers into pricing tiers based on what was carefully referred to as ‘insurance score.’ Both insurers and the credit reporting agencies developing the scoring models were quick to point out that insurance score was different than the credit score used by banks to determine if a person should qualify for loan at a certain interest rate. The ‘insurance score’ modeling vendors proved to most regulators that their scoring models had a strong independent correlation to auto insurance claim frequency. They also noted how careful they were to ensure that no other socially objectionable variables such as race, ethnicity, or income were used in developing these models. Insurance scoring also weighed various financial elements of the credit score differently than a pure credit score, to tailor the scores most appropriately for insurance use.
The results were dramatic. The early adopters of insurance scoring, companies such as Progressive, gained a significant competitive advantage by being the first companies in the market to take advantage of this ‘silver bullet’. However, despite the evidence provided by modeling firms and insurance companies that this new variable would help to more fairly price insurance customers, several state regulators were not convinced. Four states have not allowed any form of credit or insurance score to be used for insurance pricing. These states are California, Massachusetts, Michigan, and Hawaii. Recent attention to this issue has also resulted in the state of Washington placing a restriction on the use of credit for insurance pricing.
After the instant success of applying insurance scoring to the auto line of business, insurers hypothesized that the correlation of credit score to claim frequency for home insurance should be at least as strong. By the mid-2010s, virtually all home insurers had also developed a tiered pricing structure for home insurance that was based on insurance (credit) score. Many insurers assumed the same correlation between credit and claims frequency observed for the auto line of business would apply to home insurance. This can be difficult to measure precisely, and there is some evidence that the correlation may not be as pronounced for home insurance, despite many rating plans that assume a similar correlation.
There are at least two reasons for this. One is that there is a higher required credit standing related to home ownership than there is to drive a car. So, consumers seeking home insurance represent a subset of auto consumers, and as a group they are at a higher credit score level than the broader group of auto owners. Secondly, home insurance claim frequency is significantly lower per risk than auto insurance. On average, at least 1 in 10 drivers have an auto insurance claim each year, but only 1 in 30 homeowners have a home insurance claim in a given year. Many home insurance claims result from forces of nature that are beyond a person’s control, unlike auto insurance where most claims are due to at least one driver being at fault in an accident. Since the frequency of claims in home insurance is so much lower than it is for auto insurance, the assumption that credit would have a similar effect in that line takes longer to be proven than it does for auto insurance.
In addition, since a higher percentage of home claims are beyond the control of the insured, the hypothetical correlation between credit and claims frequency as it applies to home insurance loses some of its validity. Many home insurers have found that their initial pricing models overstated the impact that credit would have on underwriting results, especially for higher credit score risks.
Just how much does credit score affect rates for home insurance?
This information can be difficult to glean from the larger carriers’ publicly available rate filings, since they tend to not disclose the direct impact that credit score has on the final price. Many insurers combine credit score with other factors such as claims history in developing pricing tiers, or they mark the insurance scoring model details as confidential components of their rating plans. A recent review of both larger national carriers and smaller regional home insurers indicates that the dispersion of rate levels from high to low credit categories in most cases ranges from a factor of 1.8 to a factor of 3.0. This means that the rates paid by two homeowners with identical attributes other than credit score can be 80% to 200% (3 times) higher for a homeowner on the low end of the credit spectrum compared to that same homeowner who is in the highest credit score grouping.
At Zinsurance, our mission is to not just talk about change, but to actually change insurance for the better.
Home insurance customers are part of the group most likely to still watch network TV, and recently there is even more emphasis on television advertising targeted towards these customers. One recent campaign features two prominent NFL quarterbacks pleading for the special rate that applies only to them. Although the commercial is designed to point out that the same great rates are available to everyone, this is certainly not how auto or home insurance pricing works in practice. A high-profile NFL quarterback with a multi-million-dollar annual salary and extremely high net worth would certainly be placed into the best ‘insurance score’ tier. This would mean he would pay a rate much less than half that of a person with the same driving record and coverages, but who is on the other end of the ‘insurance score’ spectrum.
Although everyone has ‘access’ to the best rates, only a privileged few qualify for them. Proponents of insurance scoring have argued that even among high income consumers, there are those with low credit scores. Likewise, there are low-income consumers that maintain remarkably high credit scores. The same correlation between credit and claim frequency exists for these outliers as it does for the overall population. Although the statement is technically true, the point ignores the clear correlations of credit score to other attributes that should not be part of an insurance pricing model. There is certainly a strong correlation between credit score and variables that are outside a person’s control and that are clearly against public policy, such as income level, race, and ethnicity.
Who is really benefiting from the use of credit score in pricing insurance policies? To understand this better, consider that several groups have strong financial incentives to maintain the status quo. First, the credit reporting vendors that developed and maintain insurance scoring models enjoy a large revenue stream that would be taken away if the use of credit score is restricted. These vendors charge insurance companies up to several dollars per customer each time an insurance score is required in the quoting process (many quotes do not even turn into policies). This adds up to millions of dollars of revenue for these companies every month.
Who really pays this cost? Insurance companies simply pass the ‘underwriting report’ cost of obtaining these scores on to consumers in the form of higher rates. Second, insurance remains very much an ‘old school’ industry where most companies are run by executives that do not need to worry about paying bills from month to month, or whether they can make it through a job interruption. The decision makers at these companies may be out of touch with just how unfair the pricing models developed decades ago have become. Although it may be an unintended consequence, a credit-based pricing system personally benefits many insurance company decision makers in the form of lower rates for themselves and their families. Since the long-standing use of credit score allows insurance companies to more ‘accurately’ price customers, no one wants to be the first (or worse, the only) one to stand up and drive the change to a more socially equitable pricing system. This creates a strong incentive for companies to leave things as they are, to not be the first or only one to lose access to this valuable ‘silver bullet’.
If credit scoring is taken away without a replacement, consumers will pay rates based on an average credit score level. We would expect about half to end up paying more than they did before, and about half would pay less (note – most insurance company executives would pay more). This is counter to one industry contention that consumers would pay more if credit scoring were to be restricted, as a large percentage would actually benefit from lower rates.
There is one additional cost that does appear if insurance companies have the credit scoring ‘crutch’ taken away. Insurance is still one of the most inefficient and slow-to-advance financial markets. Large and small companies alike are dealing with legacy system technologies that are very costly to upgrade. The elimination of credit score as a pricing variable would further expose this weakness, and companies would need to scramble to upgrade their systems, rules, and processes to remove it. Who would pay this cost? You guessed it, consumers. Insurance companies will simply pass on the costs involved to upgrade inefficient and outdated credit-based systems on to consumers in the form of higher rate levels. However, this does not mean the industry should not do the right thing just to save consumers from a cost that is driven by its own inefficiency. The use of credit score in pricing, which has driven insurance company profits for decades, is not what is best for society in the long-term.
The ‘silver bullet’ of insurance score has meant that insurers, until now, did not need to spend resources to find better (and less costly!) ways to price customers that do not rely on the use of credit. Even most of the newer ‘insurtech’ startups have blindly incorporated credit-based insurance scoring into less-than-innovative product designs. Some of these companies are now saying they will remove credit from their pricing algorithms in the near future. Instead, they will focus on ways to differentiate risk that are based solely on individual behavior, such as the use of telematics data from mobile and smart devices. However, incorporating these less controversial (and more accurate) data elements will take time. Insurers will be hit with the double cost of removing the objectionable element of insurance (credit) score while replacing it with pricing models based on other attributes not yet captured in their pricing databases. In addition to telematics information (data gathered from mobile devices or car sensors used to track individual driving behavior), alternative methods include usage-based rating plans for auto insurance and incentives to add claim-reducing smart home devices and sensors for home insurance.
The COVID-19 pandemic has disproportionately impacted certain groups, including those skewed toward lower credit (and insurance) score bands. The events of the last several years have also led to more attention being paid to social equity and justice issues that have turned an additional spotlight on the use of credit score in insurance pricing. Some of the larger national carriers are also considering the removal of credit-based insurance score in their pricing models. The industry is also recognizing the increased pressure from regulators to find alternative pricing attributes such as telematics data and information from smart home devices. For home insurance, where comprehensive data tying individual owner behavior to claim frequency is not yet readily available, credit will not be as easy for companies to remove as a pricing variable in the near-term as it may be for auto insurance. Since the use of credit-based scoring is so firmly embedded into companies’ profit models, this transition will certainly take a long time to accomplish. Until then, many consumers will be unfairly discriminated against due to factors that are beyond their control.
At Zinsurance, our mission is to not just talk about change, but to actually change insurance for the better. In a recent survey of homeowners with higher-than-average credit scores, we found that 56% of respondents thought that credit score should not be used to price insurance. Our team has decades of insurance pricing experience and we have no doubt that credit score is predictive and has a strong independent correlation to insurance risk. However, we also recognize that it has a clear correlation to attributes that should not be used to set insurance rates. With the ever-expanding ability to access data that more accurately predicts an individual’s unique risk potential, credit scoring is a pricing attribute whose time should come to an end. The elimination of credit score as a pricing attribute, which is discriminatory and counter to required social change, is a big first step. We believe the data is now readily available to eliminate this insurance pricing ‘crutch’, and we plan to use it to create a fair insurance product and pricing system for our customers. Our product will be based solely on an individual’s experience and choices. This is the only logical path for an industry whose entire purpose is to help and protect all its customers, not just those from historically privileged groups.
Zinsurance is a D2C digital agency formed by insurance experts with a mission to change insurance for the better. We are focused on saving homeowners time and money through the use of the latest data and technology driven by an end-to-end mobile app that handles everything from quote to claim. We are innovators in all areas including marketing, coverage, product, pricing, and process. Our number one goal is to ensure our customers are protected when they need it, and we are reconstructing home insurance from the ground up to make sure we do just that.
