“Drivers of Discrimination: An Examination of Unfair Premiums, Practices, and Policies in the Auto Insurance Industry.”

Introduction

It is important to state at the outset that NAMIC is strongly opposed to broad prohibitions on the use of legitimate and predictive underwriting/rating tools. In particular, NAMIC opposes H.R. 3693, n1 the Prevent Automotive Insurance Discrimination Act (PAID Act), and H.R. 1756, n2 the Preventing Credit Score Discrimination in Auto Insurance Act. The flawed premise underlying these legislative proposals is an unsubstantiated allegation of bias and inherent unfairness arising out of the use of certain information to underwrite and rate automobile insurance. We respectfully encourage the Subcommittee to reject the legislative proposals being discussed today and to reaffirm the state-based insurance regulatory system that has served consumers so well for more than 150 years. We hope that we can demonstrate to you that additional legislation in these areas is unwarranted, unnecessary, and may unfortunately cause a great deal more harm than any alleged positive effect.

The goal of NAMIC's written statement is to provide high-level background information on the risk-based bedrock of solid insurance underwriting and pricing as well as the legal and regulatory structure surrounding insurance activity. Consistent, balanced, efficient, and effective regulation of solvency and market conduct are essential components of an insurance regulatory structure that promotes sound market principles. Within these tenets is the principle that a healthy insurance market consists of inter-industry competition, fueled by accurate underwriting and pricing. Regulatory environments governed by these guideposts facilitate consumers having a broader range of insurance products and services at fair and accurate prices while fostering financial stability. Conversely, regulatory environments that undermine these principles typically see a reduction in availability and affordability. Consumers and insurers both have an interest in a system of insurance supervision that ultimately allows a vibrant market where competitive products and services are available for consumers.

While some may have well-intentioned goals in putting forth expansive legislative proposals, we caution against falling victim to the law of unintended consequences. It has been demonstrated time and time again that such underwriting restrictions harm policyholders by driving up insurance costs across the board.

Risk-Based Rating

An insurer's main business involves providing consumers financial protection against their risks. The goal of insurance underwriting is to correlate rates for insurance policies as closely as possible with the actual cost of claims. The more accurately a company targets the actual costs, the better they are able to manage their book of business, which in turn enables them to write more products for more individuals, in more perils. Simply put, accurate underwriting enables more coverage for consumers.

For insurers, differentiating between risks in underwriting and rating is essential to meet their responsibility to pay claims and is accomplished by risk assessment, classification, and selection to achieve a sustainable portfolio at an appropriate premium. Individuals that an insurer concludes present lower risks of loss appropriately pay less per unit of insurance than individuals who present higher risks of loss. Approaches contrary to risk-based pricing - like those contained in these bills - are a direct assault on actuarial science and threaten to suppress consumer choice by doing damage to the entire auto insurance market.

Today, to comply with state insurance laws and regulations and to offer competitive rates, an auto insurance company must be able to assess risks and price policies accurately according to the likely cost of claims generated by those policies. To be clear, under state law, insurers are prohibited from setting rates that are excessive, inadequate, or unfairly discriminatory against any individual.

"Unfair discrimination" has a very specific meaning in state insurance codes, and its use is a prohibition that insurers have been subject to for many years. According to model legislation developed by the National Association of Insurance Commissioners (NAIC), there are essentially two ways in which an insurer could engage in unfair discrimination (which is explicitly prohibited): (1) making underwriting/rating distinctions "between individuals or risks of the same class and essentially the same hazard;" or (2) making underwriting/rating decisions that are unsupported by "the application of sound underwriting and actuarial principles related to actual or reasonably anticipated loss experience." n3 Reviewing these standards, it should be clear that the "unfair discrimination" standard is a term of art defining a well-established, and carefully-constructed, risk-related approach states use to directly regulate insurers. n4 This baseline standard occupies the field and state insurance departments aggressively ensure that these standards are met within their markets.

Put another way, "discriminating"/differentiating on the basis of risk is regarded by insurance experts and regulators not only as fair, but necessary. n5 In fact, pricing programs of most insurers depend on making cost-based distinctions based upon a number of different risk factors. All things being equal, the consumer who reflects a higher risk based on these factors will pay more. The fundamental goal is to match risk to rate as accurately as possible. This is done through actuarial science and consumers benefit from having as many factors as possible considered.

NAMIC seeks to preserve the ability of insurers to rely on foundational principles of actuarial soundness. The underpinning of insurance underwriting and rate-making is classifying policyholders according to predicted risk of loss by using objective and statistically supported information - the more accurate the risk assessment, the more accurate and appropriate the rate charged. Insurers make decisions based on actuarial and business principles that treat similar policyholders similarly. In other words, this process seeks to prevent unfair discrimination or capriciousness. In fact, by banning the use of risk factors that are actuarially justified - as the above referenced bills seek to do - you are requiring that insurers charge rates de-linked to risk. A lower risk individual will have to pay a higher rate just as a higher risk individual will get a lower rate. This is, at its basic, asserting an unfair discrimination standard.

Factors

Making predictions is essential - insurance differs from most other products because the actual cost of providing insurance is unknown at the time the product is offered and the customary laws of supply and demand do not apply. To most accurately make these predictions, various factors are used to analyze a risk. Looking back at historic losses helps to forecast future losses, but prior claims alone do not provide enough information to serve as an adequate predictor.

In today's auto insurance market, a multitude of risk-predicting factors may be considered, including things like: driving history, policy limits, multi-policy discounts, vehicle information, vehicle safety equipment, age, credit history, miles driven, etc. Whether and how particular factors are used differs by insurer. These underwriting and rating factors are actuarially based tools by which to assess risk and forecast loss. They are correlative, in that they are actuarially shown to have an association with loss - the two things tend to occur together. Factors do not need to demonstrate causational relationships. A causational relationship would indicate that a factor results in a loss - something that does not necessarily exist in this context. Even a fact pattern of a dozen prior speeding tickets and prior accidents does not result in a future loss. Any assertion that factors should be held to a causational standard fundamentally misunderstands the methodology of actuarial science and would exclude practically every exercise to try to assess risk.

Limiting risk factors and moving away from setting prices based on actuarially justified factors may stifle competition and may deprive consumers of the benefits that naturally flow from competition. Markets are healthier - with more options for consumers - when as many different predictive factors as possible are utilized. NAMIC members believe that competitive markets are the most effective way to ensure low prices, widespread availability of the product, superior service, and product innovation.

Furthermore, unintended consequences may result from severely restricting factors, as contemplated in these bills. In reviewing a factor, data is analyzed to determine its potential correlation with risk. A prohibition on the use of gender provides a helpful example. Data and/or studies may show that as a group, younger males may be more likely to engage in riskier driving behavior and/or that they may be more likely to have been involved with a greater number of accidents. Data from the U.S. Department of Transportation National Highway Traffic Safety Administration supports this conclusion. They found: "motor vehicle crash fatalities were higher for males than females in all age groups." n6 To void this information would seem to mean that females would pay more for auto insurance if this factor were banned.

Credit information provides another helpful example. Credit-based insurance scores now have been used by underwriters/actuaries for a few decades to more accurately assess risk and price insurance coverage. Insurance scoring provides an objective, fair, and consistent tool that insurers use with other information to better predict the likelihood of future claims and the cost of those claims. Credit-based insurance scoring is neutral on its face with respect to income, race, and ethnicity, and it is applied neutrally by insurers. The same credit standards are applied to all consumers. In other words, contrary to assertions by some of its opponents, the tool is actually an objective means of affording equal treatment in the underwriting process.

Multiple government agencies as well as others have reviewed the use of credit-based insurance scores precluded under this bill and have determined that they help insurers to better assess risk and to develop rates that are more actuarially accurate; correlate to risk of loss; and/or are not unfairly discriminatory:

o The Federal Trade Commission (FTC) study concluded that insurers' "use of credit-based insurance scores may result in benefits for consumers. For example, scores permit insurance companies to evaluate risk with greater accuracy, which may make them more willing to offer insurance to higher-risk consumers for whom they would otherwise not be able to determine appropriate premium. Scores also may make the process of granting and pricing insurance quicker and cheaper, cost savings that may be passed on to on to consumers in the form of lower premiums." n7

o The Vermont Department of Financial Regulation (DFR) found: "In general, [ ] policyholders whose premiums are based on credit-related insurance scores tend to pay lower annual premiums than policyholders whose premiums do not include insurance scores." n8

o The Nevada Insurance Division conducted a study on credit-based insurance scoring and said that their investigation "corroborates the insurance industry's contention that the majority of policyholders benefit from the use of credit scoring." n9

o The Arkansas Department of Insurance looks regularly at the use and impact of credit in personal lines. In its review of 2016 data, it determined that "80% of consumers either received a discount for credit or it had no effect on their premium." n10

o The Texas Department of Insurance (TDI) study found that "for both personal auto liability and homeowners, credit score was related to claim experience even after considering other commonly used rating variables. This means that credit score provides insurers with additional predictive information distinct from other rating variables. By using credit score, insurers can better classify and rate risks based on differences in claim experience." n11

o The Virginia State Corporation Commission Bureau of Insurance studied credit use in 1999 and again in 2016. Among the items considered, the Bureau reviewed credit-related consumer complaints and inquiries. Based on this information, it concluded that "it would be difficult to conclude that PPA [private passenger automobile] insurance policyholders are unfairly burdened by insurers' use of consumer credit information or insurance credit scores. n12

o A study out of the McCombs School of Business found: "The correlation between credit score and relative loss ratio is .95, which is extremely high and statistically significant. The lower a named insured's credit score, the higher the probability that the insured will incur losses on an automobile insurance policy..." n13

o The early EPIC Actuaries, LLC study found: "Insurance scores are among the three most important risk factors for each of the six automobile coverages studied." n14

Finally, having multiple predictive factors available for use helps to broaden the impact of any single factor. No single factor has yet been discovered which accurately measures the totality of risk represented by each insured. The state of today's actuarial science is that the most accurate risk assessment is achieved through a combination of risk factors. Every market participant understands it is in the best interest of its policyholders to have an optimal mix of factors for accurate pricing, which is also necessary to remain competitive in what is arguably the most competitive line of property/casualty insurance right now.

Market Competition Benefits Consumers

Auto insurance coverage is widely available in a highly competitive marketplace. According to S&P Global Market Intelligence, concentration in the property/casualty insurance sector as measured by the Herfindahl-Hirschman Index (HHI) decreased from 354 in 1997 to 297 in 2007. By 2017, the index increased very slightly to 301. The U.S. Department of Justice classifies any market with an HHI under 1,500 as unconcentrated and any market with an HHI over 2,500 as highly concentrated." n15 This is indicative of the strong track record and highly competitive market in the property/casualty insurance industry due in large part to business models with robust underwriting and pricing. In the twenty-year period referenced, the market has remained in a very de-concentrated range which only benefits consumers on a daily basis.

Legislation that limits underwriting factors would be a significant change that would undermine this robust competitive marketplace. Today insurers use many different rating factors and they weigh these factors differently. With a variety of factors in use, consumers can shop; they have more choice. With fewer factors, there are fewer options, more guesswork, and higher prices for consumers borne out of necessity.

The heavy level of competition ensures that insurers have every incentive to most accurately and appropriately match the rate to the actual risk that is represented by a driver. Those companies that predict claim costs better than their competitors are typically more successful. This market-driven incentive to accurately assess risk ensures that the price of insurance will be more commensurate with the level of risk that a particular policyholder presents.

Additionally, if an insurer is confident in the adequacy of its prices, there is a strong economic incentive for the insurer to provide coverage whether or not it is a high-risk applicant or a low-risk applicant. Because a range of factors have contributed to the improved accuracy of more risks, there may be stronger incentives to provide coverage to all consumers, including high-risk applicants, thereby improving availability of coverage. On the flip side, abandoning analytical techniques to more optimally assess or price risk, injects uncertainty about the adequacy of prices which may affect coverage availability. This would discard progress and return to the relatively basic and less precise methodologies that dominated a bygone era of auto insurance underwriting/rating. Reducing the factors that insurers use will unnecessarily inject uncertainty into the underwriting process affecting affordability and availability for many drivers across the country.

States Already Have Ability to Address Unfair Trade Practices

The McCarran-Ferguson Act (McCarran), approved by Congress in 1945, explicitly entrusts states with the authority and preeminent responsibility for regulating the business of insurance. Under the act, "No Act of Congress shall be construed to invalidate, impair or supersede any law enacted by any State for the purpose of regulating the business of insurance, or which imposes a fee or a tax upon such business." Under our regulatory system, property/casualty insurers across the country are subject to a comprehensive state-based regulatory and enforcement regime. Federal law or regulation that would prohibit or limit insurance underwriting factors is contrary to McCarran.

The robust state system of consumer protection regulates aspects of an insurance business such as: the company's start-up paperwork (authority and licenses), insurance products (policy forms), consistency in selection/placement (underwriting standards), product price (rate), financial responsibility (solvency), behavior (market conduct), and practices (not deceptive in claim settlement or in trade generally). This is virtually every aspect of insurance - there is no void. State departments of insurance are very active in their efforts to regulate underwriting and rating, something they are uniquely qualified to do, unlike the Federal government.

Part of consumer protection is not just ensuring that insurance premiums are not excessive, but also ensuring that they be adequate to ensure that companies have the financial ability to fulfill the promises that they make to their policyholders. Adequacy of rates requires the ability to price according to predicted losses and claim costs. When the ability to underwrite risk appropriately is dramatically impeded, financial and prudential solvency issues enter the picture. Insurers cannot on a continued basis take in less premium than they are paying out in claims or they are at the risk of potential failure. At best, a decrease in obtaining adequate premium will be a decrease in capacity to underwrite new business, which weakens the market by lessening competition.

Stringent anti-discrimination prohibitions are also in place in the states. Beyond general laws, for insurers, rating factors must be actuarially sound in order to be used by law. State regulators already have the appropriate authority to disallow any factor that they deem inappropriate or unfair. There is no data to indicate that these provisions or their enforcement have been inadequate or lacking.

It may be helpful to turn to one of the factors contemplated for prohibition, credit-based insurance scores. The vast majority of states have adopted laws or regulations to address this tool. Many are based largely on the National Conference of Insurance Legislators' (NCOIL) Model Act Regarding Use of Credit Information in Personal Insurance. n16 It was first adopted in 2002, and later amended. Among its provisions, this Model requires upfront disclosures and adverse action notices, requires prompt remedy in case of incorrect information, prohibits use of certain information, and provides "sole-basis" restrictions. Established anti-discrimination provisions apply and it explicitly prohibits insurers from using an insurance score that is calculated using income, gender, address, zip code, race, ethnicity, religion, or an individual's nationality. Additionally, this model contains language to require consideration for "extraordinary life circumstances" - that is, events such as loss of employment, death of close family member, or divorce.

Not a single court has found the use of credit-based insurance scores within this framework to be unfairly discriminatory. Indeed, the Michigan Supreme Court rejected a regulatory effort to ban the use of insurance scores, stating: "It is difficult to see how offering discounts to some insureds on the basis of good insurance scores is inconsistent with the Insurance Code's general purpose of availability and affordability of insurance for all consumers." The court noted that a rate is not unfairly discriminatory if there is a "reasonable justification" for the differential in rates "supported by a reasonable classification system" and that there was a direct, linear relationship between insurance scores and risk for automobile policies. The Court concluded that the prohibition of insurance scoring would make insurance both less available and less affordable to Michigan residents. n17

Again, general state regulatory mechanisms adequately ensure consumer protection against improper discrimination. For example, the market conduct process allows state insurance regulators to assess and ensure compliance with laws and regulations. Through this process, Departments may review complaints received by the company. At any time, a consumer complaint to an insurance department serves as an additional flag to alert regulators to possible items to investigate.

Conclusion

In conclusion, variety in objective and cost-effective underwriting tools more accurately predicts risk of loss and improves competition in the auto insurance market, which ultimately benefits consumers. Congress should take care not to sever the critical link between risk and the underwriting/rating process by unnecessarily restricting actuarially sound factors. Passing legislation to ban the use of underwriting factors would simply disrupt and substantially weaken auto insurance markets across the country, undermine the state-based system of insurance regulation that has served the U.S. well for 150 years, and ultimately harm the very consumers such action purports to help.

n1 https://www.congress.gov/bill/116th-congress/house-bill/3693?q=%7B%22search%22%3A%5B%22hr3693%22%5D%7D&s=2&r=1

n2 https://www.congress.gov/bill/116th-congress/house-bill/1756?q=%7B%22search%22%3A%5B%22hr1756%22%5D%7D&s=3&r=1

n3 "Unfair Discrimination," Sec. G(3), Unfair Trade Practices Act, NAIC Model Regulation Service, January 1993, pp. 880-884. MDL-880.

n4 NAMIC has explained the term "unfair discrimination" in the context of insurance regulation previously and will do so in greater detail below. Before turning to that background, it may be helpful to consider what is not intended. While used as a term of art in laws, regulations, actuarial standards, and articles for years, the wording seems unfortunate because those from outside the industry may default to interpreting activity in the context of the everyday meaning of "discrimination," a practice that is repugnant and contrary to our notions of fairness. The variance between the everyday meaning and the insurance standard begins with a working definition of "discrimination." The everyday meaning could be considered "prejudicial treatment of people, especially on the grounds of income, race, national origin, etc." The insurance meaning is more akin to a secondary definition of distinguishing and could be considered "recognition and understanding of the difference between one risk and another without consideration of the categories [above]." NAMIC highlights this not to diverge from the long-established definitions of "unfair discrimination," but to acknowledge that the use of this meaning could pose some challenge in ensuring a common vocabulary in a contemporary dialog of these issues outside of the industry's functional regulator.

n5 A leading textbook for students of insurance regulation instructs that "in insurance, discrimination is not necessarily a negative term so much as a descriptive one. For insurance, fair discrimination is not only permitted, but necessary." See Kathleen Heald Ettlinger. State Insurance Regulation (Insurance Institute of America, 1995), pp. 29-30.

n6 https://crashstats.nhtsa.dot.gov/Api/Public/ViewPublication/810853

n7 Federal Trade Commission: Report to Congress on Impacts on Consumers of Automobile Insurance (2007): https://www.ftc.gov/sites/default/files/documents/reports/credit-based-insurance-scores-impacts-consumers-automobile-insurance-report- congress-federal-trade/p044804facta_report_credit-based_insurance_scores.pdf

n8 Vermont Department of Financial Regulation: Study of Credit-Based Insurance Scoring (2016).

n9 "Report on the Use of Consumer Credit and Loss Underwriting Systems," Nevada Dept. of Business & Industry, Division of Insurance (2005).

n10 Report to the legislature, "Use and Impact in Personal Lines Premiums Pursuant to Ark. Code An. Sec.23-67-415" (2017): https://insurance.arkansas.gov/uploads/resource/documents/2017credit.pdf

n11 Texas Department of Insurance, "Supplemental Report to the 79th Legislature: Use of Credit Information by Insurers in Texas: The Multivariate Analysis" (2005): https://www.tdi.texas.gov/reports/documents/credit05sup.pdf / https://www.tdi.texas.gov/reports/credit3.html

n12 State Corporation Commission's "Use by Insurers of an Insured's or Applicant's Credit Information in Connection with Underwriting Motor Vehicle Insurance Policies" (2016): https://rga.lis.virginia.gov/Published/2016/RD331/PDF

n13 See "A Statistical Analysis of the Relationship Between Credit History and Insurance Losses," University of Texas Bureau of Business Research at the McCombs School of Business (2003).

n14 "The Relationship of Credit-Based Insurance Scores to Private Passenger Automobile Insurance Loss Propensity" (2003): https://www.ask-epic.com/Publications/Relationship%20of%20Credit%20Scores_062003.pdf

n15 https://www.iii.org/sites/default/files/docs/pdf/insurance_factbook_2019.pdf

n16 http://ncoil.org/wp-content/uploads/2016/04/11262015PropertyCasualtyModelAct.pdf

n17 Insurance Institute of Michigan v. Commissioner, 785 N.W. 2d 67, at 83 (Mich. 2010).

Read this original document at: https://democrats-financialservices.house.gov/UploadedFiles/HHRG-116-BA04-Wstate-CollinsE-20200304.pdf