The laws that dictate what an insurer can charge and deny

Rate filing: why insurers can't just charge whatever they want

Most US states require insurers to file proposed rates with the state regulator before (or shortly after) using them. This is called rate filing, and it comes in a few flavors:

• Prior approval: the insurer must get the regulator's sign-off before using a new rate. Common for auto and home insurance in states like California.
• File-and-use: the insurer files the rate and can use it immediately, but the regulator can later reject it.
• Use-and-file: the insurer uses the rate first and files shortly after.
• Flex rating: rates within a set band (say, plus or minus 5 percent) go through with minimal review; anything outside the band needs approval.

The regulator's job in each case is to check that rates are, per the standard language found in most state statutes, "not excessive, not inadequate, and not unfairly discriminatory." That three-part test, first popularized through NAIC model laws, is the backbone of US rate regulation:

• *Not excessive*: the insurer isn't gouging policyholders relative to the risk and expected payout.
• *Not inadequate*: the insurer isn't underpricing to the point of threatening solvency (a risk to everyone else in the pool if the company can't pay claims).
• *Not unfairly discriminatory*: similarly situated policyholders should be charged similarly, and any differences must be actuarially justified.

That last clause is where the real fight happens.

"Unfairly discriminatory" versus discrimination law

Here's the conceptual trap: "unfairly discriminatory" in insurance rate law does not mean the same thing as discrimination under civil rights law.

Insurance pricing is *supposed* to discriminate, in the statistical sense: a 68-year-old smoker and a 25-year-old non-smoker should not pay the same life insurance premium. Actuarial classification is the business model. The legal question is which classification variables are permissible and which cross into illegal territory.

Unfair Trade Practices Acts (most states have adopted a version of the NAIC's Model Unfair Trade Practices Act) prohibit specific conduct, including:

• Misrepresenting policy terms
• Unfair claims settlement practices (lowballing, unreasonable delay)
• Rebating (illegally giving kickbacks to attract business, prohibited in most states)
• Discriminatory pricing or underwriting on prohibited bases

The prohibited bases vary by state and by line of business, but commonly include race, religion, national origin, and sometimes gender, genetic information, or credit history for certain products. The federal Genetic Information Nondiscrimination Act (GINA, 2008) bars health insurers and employers from using genetic test results in underwriting. Many states go further and restrict its use in life and disability insurance too.

Credit-based insurance scores are a good case study of the gray zone. They're legal to use for auto and home pricing in most states, and insurers argue they're statistically predictive of claims. But several states (California, Massachusetts, Hawaii) ban or restrict their use, partly because of proxy discrimination concerns: credit scores correlate with race and income even without any intent to discriminate. This is the modern frontier of the "unfairly discriminatory" test, and regulators are increasingly using disparate impact analysis, not just intent, to evaluate it. The NAIC has an active working group on this; see the NAIC's overview of big data and AI in insurance for ongoing regulatory thinking.

What this means for denial, not just pricing

The same statutes that govern pricing govern denial. An insurer generally cannot refuse to issue a policy based on a prohibited characteristic, and it must have an actuarially sound, filed rationale for any risk-based decline.

This shows up concretely in:

• Health insurance: the Affordable Care Act (ACA, 2010) banned denial or rate-up based on pre-existing conditions in the individual and small group markets, a massive change from pre-2010 underwriting norms.
• Life and disability insurance: insurers can still medically underwrite, but state insurance codes typically require that any decline be based on demonstrable risk factors, not blanket exclusions of protected classes.
• Property insurance: practices like historical "redlining" (denying coverage by neighborhood) are illegal under both state unfair trade practices law and, where housing-related, the federal Fair Housing Act (1968).

How a filing actually gets challenged

Picture a workers' compensation insurer filing a 12 percent rate increase in a prior-approval state. The process typically looks like this:

1. The insurer submits an actuarial memorandum: loss history, trend assumptions, expense loads, and a proposed rate.

2. The state's rate analysts (often actuaries themselves) review the filing against the excessive/inadequate/discriminatory test.

3. The regulator may request a public hearing, especially for large increases on personal lines like auto or homeowners.

4. Consumer advocates or intervenors (in some states, groups can formally challenge filings, California's Proposition 103 (1988) explicitly created this right) can dispute the assumptions.

5. The regulator approves, modifies, or rejects the filing.

This is slow and adversarial by design. It's also why insurers in heavily regulated states (California is the classic example) sometimes exit or restrict writing new business rather than fight prolonged rate suppression, a real dynamic that played out with several homeowners insurers in California in 2023 to 2024 amid wildfire risk repricing disputes.

🎬 [VIDEO: "How Insurance Rate Regulation Works" - https://www.youtube.com/results?search_query=how+insurance+rate+regulation+works - search result page surveying explainer videos on state rate filing and approval processes]

The compliance function's job

For a compliance officer inside an insurer, this legal architecture translates into concrete daily work:

• Maintaining a filed rate manual and ensuring no agent or system charges outside the filed rate (a real violation even if the off-filing rate is *lower*)
• Auditing underwriting models (especially AI-driven ones) for proxy discrimination before deployment
• Training claims staff on unfair claims settlement practice rules to avoid state market conduct exams, in which regulators audit a sample of files for compliance patterns
• Tracking state-by-state variation, since a rating factor legal in Texas may be banned in California

Key Takeaways

• US insurance rate regulation happens state by state under the McCarran-Ferguson Act, with the NAIC providing model laws but no direct enforcement.
• Every filed rate must pass the "not excessive, not inadequate, not unfairly discriminatory" test; that third prong is the legal battleground over pricing variables like credit scores and ZIP code.
• Unfair Trade Practices Acts govern both pricing and denial conduct, banning things like misrepresentation, rebating, and discrimination on prohibited characteristics, which vary by state and product line.
• Landmark federal laws layer on top of state rules: GINA restricts genetic data use, the ACA banned pre-existing condition underwriting in health insurance, and the Fair Housing Act bars discriminatory property coverage decisions.
• Proxy discrimination, where a facially neutral variable correlates with a protected class, is the active regulatory frontier, especially as insurers adopt AI-driven underwriting models.