What Insurance Is Actually For, and When to Skip It

Every insurance policy is the same transaction: you pay a known, small, certain cost to avoid an unknown, large, uncertain one. The insurer takes on your risk, pools it with thousands of other people's, and prices it so that on average they collect more than they pay out.

That last part is important, and it is not a criticism. It is how the model has to work. The consequence is that on average, across everyone, insurance is a losing bet by design. You are not buying an investment. You are buying protection from an outcome you could not otherwise absorb.

Once you see it that way, the decision rule becomes clear.

The rule: insure what you could not survive

Insurance is worth its cost when the potential loss would be financially catastrophic - when it would wipe out your savings, force you into debt you could not repay, or eliminate your household's income.

It is poor value when the potential loss is one you could absorb from savings without changing your life. In that case you are paying the insurer's margin for the privilege of avoiding an inconvenience.

This is why the same product can be essential for one household and wasteful for another. The variable is not the risk. It is your ability to absorb it.

Where this points

Usually worth it:

• Health cover, because medical costs have effectively no ceiling
• Liability cover, for the same reason - a judgment against you is not capped by what you own
• Home or property cover, where the asset is a large share of your net worth and you could not replace it from cash
• Income protection or life cover, where other people depend on your earnings

Usually not worth it:

• Extended warranties on consumer electronics, where the maximum loss is the price of the item
• Low-value item cover, phone insurance being the common example, where the annual premium plus excess often approaches replacement cost
• Cover that duplicates something you already hold, which is more common than people realise - credit cards and home policies frequently include travel or item cover you are separately paying for
• Life cover with no dependants, where there is no income stream to replace

Two things people get backwards

Buying more cover for small risks and less for large ones. It is psychologically easier to insure a $900 phone than to think about disability cover, but the phone is the loss you could survive. Attention and premium budget should follow severity, not familiarity.

Treating deductibles as something to minimise. A low deductible is not better protection - it is the same protection with more of the small losses shifted back to the insurer, at a price. If you can cover the deductible from savings, taking a higher one and keeping the difference is usually the stronger position. We go through the arithmetic in choosing an insurance deductible.

Self-insuring is a real option, with a condition

Choosing not to insure a risk is called self-insuring, and it is entirely legitimate - provided you actually set the money aside. Skipping a $300 annual premium and spending the $300 is not self-insuring. It is just being uninsured.

The honest version is to hold the reserve. If you would not fund the reserve, buy the policy.

Before you buy anything

• Check what you already have. Employer benefits, credit card terms and existing policies frequently overlap
• Read what is excluded, which tells you more about a policy than what is covered
• Confirm the claims process and what documentation you would need
• Check the insurer is licensed in your state through your state insurance department

The summary

Insurance is a tool for one job: preventing a survivable event from becoming an unsurvivable one. Judged against that job, most households are underinsured on the few risks that could ruin them and overinsured on the many that could not.

Sorting risks by the only question that matters

The test is not how likely a loss is, or how unpleasant. It is whether you could absorb it. Two dimensions, four quadrants:

Almost every good insurance decision sits in the bottom-right box: events unlikely enough to be cheap to cover, severe enough that you could not survive them from savings.

Almost every bad one sits in the top-left, where the loss is small enough to pay for and frequent enough that the insurer must charge more than it costs you on average.

Why insuring small risks costs more than it returns

An insurer must collect more in premiums than it pays in claims, plus administration, commission and profit. That margin is the price of transferring risk, and it is entirely rational.

It also means that across many small, affordable risks, you will pay more in premiums than you receive in claims. That is not a flaw or a scam; it is arithmetic. The margin is worth paying when the alternative is a loss you could not survive, and not worth paying when the alternative is an inconvenience.

Extended warranties are the clearest case. They cover a moderately likely event with a modest cost, sold at a high margin, frequently by the retailer who benefits from the sale. The expected value is poor and the loss is affordable — both signals point the same way.

The two people get backwards

Under-insuring catastrophe. Disability insurance is the most under-held major cover. Your ability to earn is usually your largest asset, and losing it is both plausible and financially unsurvivable, yet it is insured far less often than a phone.

Liability is the second. It costs relatively little to raise liability limits, and it covers exactly the kind of event — being responsible for serious injury to another person — that has no ceiling and no savings-based solution.

Over-insuring inconvenience. Extended warranties, low deductibles chosen out of habit, credit card payment protection, and cover for individual items that would be annoying rather than ruinous to replace.

The pattern is understandable: small risks feel real because they happen, while catastrophic ones feel abstract because they mostly do not. The arithmetic runs the other way.

Self-insuring, and its one condition

Choosing not to insure a risk is a legitimate decision — but only if you have actually set aside the money. Self-insurance means holding the reserve, not simply declining the cover and hoping.

Declining an extended warranty and putting nothing aside is not self-insuring. It is being uninsured and calling it a strategy. The distinction shows up at the moment the appliance fails.

This is why deductible choices and cash reserves are the same decision, and why building an emergency fund sits underneath every sentence in this article. Every policy you hold leaves a deductible-sized gap you have agreed to cover, and self-insured risks sit on top of that.

Before you buy anything

1. Name the loss. What specific event does this cover, and what does it pay?
2. Ask what happens without it. If the answer is "I would pay for it", you have your answer.
3. Check whether you already have it. Cover duplicates constantly — travel insurance included with a credit card, liability under a homeowners policy, life cover through an employer.
4. Read the exclusions before the benefits. The exclusions determine whether the cover applies in the situation you are actually worried about.
5. Prefer raising limits over adding policies. Increasing liability limits on cover you already hold is usually far cheaper per dollar of protection than a new policy.
6. Check the deductible you would face, and whether you hold it in cash. See choosing an insurance deductible.

The summary

Insurance is not a savings vehicle, an investment, or a way to avoid inconvenience. It is a mechanism for converting a loss you could not survive into a premium you can budget for.

Judged that way, most people should hold fewer policies with higher limits: comprehensive cover on the handful of risks that could genuinely end their financial life, and none at all on the many that would merely annoy them. The money saved on the second category usually pays for a meaningful improvement in the first.

Review it when your life changes, not annually

Cover tends to be set once and left, which means it drifts out of alignment with what it is supposed to protect.

The events that should trigger a review are the ones that change what you could not survive:

• A child arriving, which creates a dependant and usually a large life insurance need
• A mortgage, which adds a debt that would pass to your household
• A significant pay rise, since income replacement cover sized years ago is now too small
• A child becoming independent, which may allow cover to be reduced rather than increased
• Starting a business, which frequently removes employer cover and adds liability exposure
• Paying off the mortgage, which reduces the need

Reviewing on a calendar tends to produce either no change or an upsell. Reviewing on these events produces changes that track the actual risk.

Sources

• NAIC: How does insurance work?
• USA.gov: Health insurance

Current as of August 2026. Insurance is regulated at state level; availability and terms vary. This is general information, not a recommendation about any specific policy.