Insurance Misconceptions: Why the Cost-Cutting Intuitions Fail

Corrections to the intuitions that most often mislead people about insurance, kept out of the concept chunks. ## About price **Premiums are not mostly profit.** In the US property and casualty industry's best underwriting year in over a decade, the combined ratio was about 96.5 — roughly 72 cents of each dollar to claims, 25 to expenses, 3 to profit. Homeowners specifically ran about 99.7. Removing the profit entirely changes the price by a few per cent. See Combined Ratio: Where an Insurance Premium Actually Goes. **"Insurance is expensive" usually means "risk is expensive."** The premium tracks expected loss. A high premium is information about the property, not about the insurer. **A thin margin is not a sign of a broken market.** It is what competition looks like after it has worked. ## About pooling **More members does not mean cheaper per member.** Each member brings claims as well as premium. Pooling converts an unpredictable large loss into a predictable small payment; it does not reduce the total. See Risk Pooling Reduces Uncertainty, Not Cost. **Bigger is not automatically safer.** Diversification requires *independent* risks. A hundred thousand houses in one neighbourhood is a far worse pool than ten thousand spread across a continent — a fact that inverts the intuition that a global insurer would be the most robust. **The gains from scale are exhausted early.** The benefit decays as 1/√n, so almost all of it is captured by the time a pool reaches regional or national size. ## About personal experience **"I've never claimed, so I'm being ripped off" and "the pool sees thousands of claims" are the same fact.** At roughly a 1-in-18 annual claim rate, any individual year is overwhelmingly uneventful — and 5.5% of a hundred thousand houses is still thousands of claims. Personal experience is calibrated on a few dozen house-years; the insurer sees millions. Over thirty years, most homeowners do claim at least once. **Premiums from people who never claim are not idle money.** They are the mechanism that pays the people who do. There is no reservoir of unused premium — about 72 cents of every dollar has already flowed out as someone else's claim. Refunding non-claimants would remove the funding for claims. **What you bought was the protection, not a claim.** Thirty years of not needing it is the good outcome. ## About the structures **Risk-based pricing exists because flat pricing is unstable, not because insurers are discriminating for profit.** Any insurer pricing flat is picked apart by a competitor who segments. See Risk-Based Pricing and the Adverse Selection Spiral. **A member-owned non-profit insurer is not a thought experiment.** Mutuals have run that model for a century and land 10–30% cheaper, not free. See Mutual Insurance and the Assessable Mutual. **"Just chip in when someone has a loss" was the original model and was abandoned.** Assessments cannot be collected after the benefit has been delivered, and the real arithmetic is far above the $5–10 per month the intuition suggests. It survives only in small socially-bound communities. **Insurance is not unregulated.** Capital is ring-fenced per jurisdiction under regimes like Solvency II and US risk-based capital rules — which is a separate, independent reason a single global insurer could not exist. **Government insurance is not automatically cheaper or automatically worse.** Switzerland's cantonal monopolies are cheaper than private cover; the NFIP owes over $22 billion to the Treasury. The variable is whether prices are allowed to track risk. See Public Insurance: Where Governments Already Nationalised Risk. ## About the numbers Figures in this area vary sharply with methodology, and quoted averages should be treated with care. Published Florida premium averages differ by a factor of two or more depending on whether they use regulator filings for a standard policy or broader market surveys with different coverage assumptions. Claim frequency and severity averages likewise shift with the line of business and the year. Treat any single headline figure as an order-of-magnitude indicator.