The insurance business uses underwriting rules to decide who will be insured and at what cost. Insurance is based on risk; thus, insurance firms establish specific rules to determine when they will accept that risk and when they will not because a person is too dangerous to cover. Although insurance firms are barred from discriminating based on characteristics such as race, some information about individuals is used to assess risk and determine prices. This indicates that discrimination in some form is both essential and permissible. Still, discussions about what is fair or unfair discrimination are getting more and more attention, especially after the murder of George Floyd in 2020.
Actual Discrimination
The industry’s standard-setting body is the National Association of Insurance Commissioners (NAIC). In response to the George Floyd demonstrations, the NAIC convened a special session on race to examine the relationship between insurance and racial prejudice. While overt racial discrimination has become less prevalent, NAIC members assert that other types of intolerance exist, particularly in using big data. In addition, as mentioned below, lawsuits and investigations have argued that long-standing discriminatory practices, such as redlining and premiums based on race, continue to harm the sector.
The provided categories need to be more comprehensive. For instance, health insurance has been a source of worry, especially with federal regulations. For example, on June 19, 2020, the U.S. Centers for Medicare and Medicaid Services issued a final Nondiscrimination in Health and Health Education Programs or Activities decree.
The decision was quickly attacked by California Insurance Commissioner Ricardo Lara, among others, as a barrier to healthcare access for LGBTQ+ individuals, those with disabilities, and anybody whose first language is not English.
Variations on Discrimination
Guidelines for insurance underwriting use a type of discrimination based on risk profiles. To compute premiums, they divide individuals into high- and low-risk groups and push clients to lessen their dangerous habits. This is deemed permissible, although the history of underwriting is rife with unacceptable prejudice, also known as unjust discrimination. Under U.S. law, underwriting standards are prohibited from using unjust discrimination. Unfair discrimination targets protected classifications, including race, national origin, sexual orientation, and religion—forms of discrimination range from increased premiums and weakened plans to denial of coverage.
DISPARATE IMPACT VS. UNFAIR DISCRIMINATION
According to Susan T. Stead, Esq., discussions concerning algorithmic modeling in insurance tend to combine unjust discrimination and disproportionate effect, which are legally distinct concepts. According to Stead, disparate impact is a legal strategy for proving discrimination against a protected class without “overt discrimination.” In contrast, unjust discrimination occurs when identical risks are addressed differently for reasons unrelated to risk. It is prohibited by law in all states.
The University of Michigan Law School concluded in 2013 that anti-discrimination rules “vary significantly” by state and insurance type. A “surprising” number of counties lacked explicit legislation against unfair discrimination based on race, indicating that the federal government must play a more vital role in regulating race-based discrimination in insurance.
Examples of Important Underwriting Guidelines Discrimination
Housing and Redlining
Redlining is a kind of discrimination that has garnered widespread attention in recent years due to its enduring impact on inequality. The tradition dates back to the administration of President Franklin Delano Roosevelt. Around this period, the federal government guaranteed house mortgages to increase homeownership and the White middle class. The Home Owners’ Loan Corp, a government agency, categorized neighborhoods across the United States based on a perceived degree of risk based on factors such as: • Age and condition of the housing; • ability to have transportation • Community amenities • Proximity to undesirable properties (e.g., polluting industries) • Residents’ employment status and economic class • Residents’ ethnic and racial composition
On maps, communities were color-coded based on danger. Red denote predominantly minority ethnic and racial communities (thus “redlining”). Some regions were deemed “dangerous,” hence lenders declined to provide credit. Redlining redirected resources, such as loans and insurance, away from communities of color.
The Federal Housing Administration’s (FHA) 1938 Underwriting Handbook detailed the blatantly racist nature of these maps. Under a section titled “Quality of Neighborhood Development,” it even included “incompatible ethnic and social groupings” and the “chance of the place being occupied by such groups” as underwriting negatives, along with “flimsy building” and “bizarre architectural styles.” “For a neighborhood to remain stable, the same socioeconomic and racial classes must continue to occupy the same homes. The guideline said that a change in social or racial occupancy often led to instability and reduced values. The destructive impacts of these maps and racially discriminatory covenants on real estate prices and generational wealth continue to this day.
Act for Civil Rights
Since then, more overt kinds of discrimination have been prohibited. In the 1948 decision Shelley v. Kraemer, the U.S. Supreme Court determined that racial covenants are invalid because they violate the 14th Amendment. Significantly, the Civil Rights Act of 1964 prohibited racial discrimination. This had an effect on life insurance rates depending on race, as detailed below.
Numerous additional projects in this region would explicitly include redlining after the killing of the Rev. Dr. Martin Luther King Jr. the Fair Housing Act of 1968 prohibited redlining based on race.
The Housing and Urban Development Act of 1965, intended to integrate federal housing initiatives, introduced grants for low-income homeowners, rent subsidies for the aged and physically challenged, better access to public housing, and advantageous loans for war veterans. In addition, the Home Mortgage Transparency Act of 1975 requires lenders to report census data related to their loans (HMDA).
Despite this, it is said that this prejudice continues to exist in practice. For instance, a series of New York cases asserted that redlining tactics remained into the twenty-first century.
Race and Existence Assurance
According to an essay published in the Northwestern Journal of Law & Social Policy by Mary L. Heen, the United States’ life insurance market has a history of maintaining racial inequalities. She believes that after Reconstruction, the insurance business used high death rates and intrinsic racial disparities to justify offering formerly enslaved people just two-thirds of the benefits granted to White people. Companies with race-based premiums tended to disregard information that did not conform to preconceived hierarchies, such as women having a lower death rate, indicating that risk was not the paramount consideration in determining premium rates.
Similar behaviors persisted until the twenty-first century. In 1940, for example, the NAIC produced a report examining death rates by race, which insurers subsequently used to establish premiums based on race. According to the NAIC, the research fuelled discriminatory underwriting procedures until some time after the use of race was prohibited. Insurers kept two sets of rate books at the time, with one showing higher rates for Blacks, who mainly purchased “industrial life insurance” to pay funeral expenses. The coverage given to African-Americans was less comprehensive and costly, with premiums as much as 30 percent higher.
Race-based premiums remained lawful until 1964, under the administration of Lyndon B. Johnson, when the Civil Rights Act was passed in response to demands from civil rights campaigners.
Race and Vehicle Insurance
In 1897, the first auto insurance coverage surfaced in the United States. New Hampshire was the first state to establish a state insurance legislation requiring insurers to provide particular types of coverage, known as an assigned risk plan, in 1938. When Massachusetts adopted no-fault insurance in 1970, it was a later development. In the 1970s, guaranteed access to motor insurance would also become available. According to the NAIC, South Carolina approved a statute in 1976 ensuring vehicle insurance coverage to all qualified residents within its territory. Other 1970s reforms addressed access to motor insurance. Among the significant things from that decade and the next are:
• In 1977, a state study from the Michigan Insurance Bureau advocated subjecting vehicle insurance underwriting criteria to bureau control to eliminate “subjectivity.”
• In 1978, Massachusetts enacted a statewide system for regulating vehicle insurance that guaranteed access and prohibited the use of protected factors in determining premiums.
• The Michigan Supreme Court ruled in 1978 that no-fault coverage rules were unconstitutional.
• In 1986, the Government Accountability Office (GAO) performed a detailed analysis of vehicle insurance, including how states influenced the cost and availability of insurance that limited the variables that insurers use to price their unlawful products.
In recent years, studies have found that redlining has persisted in the car insurance sector in various ways. An examination released in 2017 by Consumer Reports and ProPublica using payment data identified disparities in vehicle insurance rates in California, Illinois, Missouri, and Texas that cannot be explained by variations in risk, indicating a “subtle kind of redlining,” in the words of the authors. Doug Heller, an insurance expert for the CFA, stated, “The companies will insist that they never ask a customer’s race, but if they are serious about confronting systemic racism, they must acknowledge that their pricing tools use proxies for race that make government-mandated auto insurance more expensive for Black Americans.”
Such bills as H.R. 3693 and H.R. 1756 have been introduced in the U.S. House to restrict discriminatory vehicle insurance practices. These two measures from 2019 aimed to limit the use of income proxies and credit scores to determine insurance premiums, but the full House or Senate never passed them.
In 1995, Fair Isaac Inc. (later FICO) and ChoicePoint were the first to implement credit ratings into vehicle insurance. Opponents have suggested it is a “substitute for redlining” and raises premium prices for minority neighborhoods.
Colorado approved a law in 2021 that prohibits discrimination against various classifications, including race, sexual orientation, and gender identity and expression. The rule compels insurance firms to show that their “use of external data and complex algorithms does not discriminate based on certain classifications, including ethnicity, sex, sexual orientation, and gender identity.” With this new regulation, insurance firms will be required to demonstrate that their pricing schemes do not discriminate against otherwise excellent drivers, according to Consumer Reports’ Chuck Bell. “Colorado now has the tools necessary to prevent discrimination and guarantee that vehicle insurance is priced fairly so everyone can afford the necessary coverage.”
Using Algorithms
Algorithms are the instructions insurance firms use to determine premiums (and trade stocks and manage asset liability, among other uses). Yet, algorithms may lead to prejudice in the underwriting of insurance policies. As part of a modernization project funded by Congress, the FHA unveiled an algorithmic underwriting system for single-family forward mortgages in 2020. It was the first system to be created by the FHA, which aimed to simplify the mortgage process.
Nonetheless, concerns remain over the actual effect of algorithmic insurance procedures. Proponents assert these algorithms foster or amplify prejudice, prompting governmental recommendations to address the problem. The 2020 Data Accountability and Transparency Act would have established a federal agency to safeguard the privacy and prohibit the use of personal data for discrimination against protected groups.
It also zoomed down on underwriting methods and would have necessitated regular testing of such systems for bias. According to experts reading the law, when prejudice was discovered, it would have been essential to demonstrate that the algorithm was necessary, that its purpose could not have been fulfilled in nondiscriminatory ways, and that the discrimination was not deliberate. The Diversity and Inclusion Data Accountability and Transparency Act was reintroduced in the House of Representatives in March 2021.
There are additional restrictions on algorithmic techniques at the state level. New York, for instance, prohibits insurers from using algorithms that “would have a discriminatory effect on the protected populations listed by New York and federal law.” Nevertheless, regulatory experts have argued that insurers in the state need help to gather information on legally protected classifications, making it difficult to determine the impact of algorithms and considerably confounding this requirement. Other states, including California, Connecticut, Illinois, Maryland, Massachusetts, Michigan, and New Jersey, have enacted or considered restrictions on the inclusion of personal information in underwriting, ranging from limitations on the use of genetic data in life insurance to the consideration of education, employment, and ZIP code criteria.
What criteria do insurance providers evaluate?
The specifications vary from business to company and insurance product to insurance product. Yet, insurance underwriters look for risk variables outlined by their company’s underwriting rules. For instance, life insurance underwriters consider age, gender, health history, marital status, and smoking/drinking habits. On the other hand, auto insurers evaluate driving records, age, gender, years of driving experience, and claim history.
What Constitutes Unfair Insurance Discrimination?
Discrimination occurs when equivalent risks are addressed differently, and premiums are not based on proportional risk but on characteristics like color.
What Does Redlining Entail?
Redlining is the discriminatory practice of rejecting loans or insurance to inhabitants of specified regions based on race or ethnicity. This practice is now banned. Sociologist John McKnight invented the word in the 1960s to describe color-coded maps prepared by the Home Owners’ Loan Corp. (HOLC) that labeled districts inhabited by ethnic minorities as “dangerous” to lenders and highlighted them in red. Redlining led mainly to the current racial wealth inequality.
The Bottom Line
NAIC members who presented at the 2020 session on race recommended several remedies for existing inequalities, including increasing minority representation in the industry, educating consumers, and regulating big data to ensure transparency, protect privacy, and discourage discrimination. In addition, the NAIC has created a special committee to address these challenges.
