What is insurance, and when is it worth buying?
It is worth buying when the loss would be one you could not absorb, and the reason sits in how the price is set. An insurer collects premiums from many people, pays claims to the few who suffer losses, and sets the premium high enough to cover the claims it expects plus its own costs of doing business. That is risk pooling, and it works: it converts an unpredictable catastrophe into a predictable line in your budget. But it also means that across all policyholders, more money goes in than comes back out. Insurance is not a way to come out ahead on average. It is a way to make sure a bad year does not become a permanent setback, and you pay a margin for that certainty.
One implication follows from that, and it drives every decision on this page: paying a margin makes sense when the alternative is ruin, and makes no sense when the alternative is an inconvenient month. That single test explains why liability coverage is usually under-bought and appliance warranties are usually over-bought.
The vocabulary every policy uses
A handful of words do most of the work, and they are worth pinning down because every later section uses them.
- Premium is what you pay to hold the policy, whether or not you ever claim. It buys the coverage and nothing else: it is not credited toward your deductible and it does not count toward any of the ceilings below.
- Deductible is what you pay out of your own pocket before the insurer starts paying.
- Copay and coinsurance are your share of a covered service. A copay is a flat amount per visit or prescription; coinsurance is a percentage of the plan's allowed amount, the negotiated price, rather than of the provider's billed charge. Coinsurance generally applies only once the deductible is met; whether a copay does is a plan design question, and many plans charge one from the first day of the year.
- Out-of-pocket maximum is an annual ceiling on your own share, after which the plan pays everything. Health plans have one. Most property policies do not, which is a difference readers routinely carry from one to the other by mistake.
- Policy limit is the ceiling on what the insurer pays. In a genuine catastrophe it is the only number that matters, and it is the one almost nobody checks. An exclusion is something the policy does not cover at all, at any limit.
Two of those are levers rather than costs. Raising a deductible lowers the premium, which trades a certain annual expense for an occasional larger one. That is the insurance transaction run in reverse, and it is a good trade only if you can actually pay the deductible on short notice, which ties every policy you own to the cash you keep on hand. Raising a policy limit usually costs surprisingly little, because the claims that reach a high limit are rare.
One more mechanic explains a lot of frustration: underwriting. Insurers price each applicant on their own record, health, and history, and they can decline. That is why the cheapest time to buy most coverage is before you need it, and why letting a policy lapse in a tight month can be irreversible in a way that cancelling a subscription is not.
Finally, buying a policy is only one of four things you can do with a risk. You can reduce it, by installing a smoke detector or fencing a pool. You can avoid it, by not owning the boat. You can transfer it to an insurer. Or you can knowingly retain it, paying for the loss yourself if it happens. Most households need some of all four, so the useful question is never just "should I insure this" but "how should I handle this".
Which risks actually need insuring?
The ones that are severe enough to change your life and unlikely enough to be affordable to cover. Three questions, asked in order, sort almost everything: how large is the worst case in dollars, how likely is it, and could your household absorb it without derailing a goal. A large, unlikely, unabsorbable loss is precisely what insurance prices well. A small loss is handled more cheaply by cash than by any policy. And a loss that is both large and likely tends to be either uninsurable or priced so close to the loss itself that the policy stops being a bargain.
Run that test across a household and the same short list of genuinely catastrophic risks comes up every time.
- Dying while other people depend on your income. The loss is every future paycheck those people were counting on.
- Losing the ability to earn. For most households with working years still ahead, the income yet to be earned is the largest asset they have, by some margin over the house. It is what pays for everything else on this list, and it is routinely the only major asset left uninsured, while the house is insured without a second thought.
- A serious illness or injury. Medical bills are the loss most likely to arrive regardless of how careful you are.
- Being held responsible for someone else's injuries. This is the only risk on the list with no natural ceiling, because the cost is another person's medical care and lost earnings rather than the value of anything you own.
- Losing a home or a car. Large, and capped at what the thing is worth.
- Needing years of help with daily living. The risk most people discover is uncovered at the worst possible moment.
- A business you own being sued, or losing you. Applies to anyone self-employed, which is a larger share of readers than it used to be.
- Outliving your money. The odd one out, because the loss is not an event but a duration. It is the one risk on this list transferred with an annuity rather than a policy: you hand an insurer a sum and it promises income for as long as you live, which is what makes longevity risk insurable at all. The products and the trade-offs belong to retirement planning, and this page hands off to them there.
Job loss deserves a mention precisely because it is the exception. State unemployment insurance replaces part of the income for a limited stretch, with most states capping regular benefits at 26 weeks. Beyond that there is no meaningful standalone private market for job-loss coverage in the United States. What does exist is sold as an add-on to a specific debt, covering the payments on an auto loan or a credit card if you are laid off, and consumer regulators have flagged those products as expensive relative to what they pay out. So the rest of the answer here is cash set aside in advance rather than coverage purchased.
What gets under-bought, and what gets over-bought
The pattern is consistent, and the reasons are psychological rather than financial. Coverage gets under-bought when the risk is hard to picture. Disability feels remote in a way that a car accident does not, and the policy at work feels like the matter is handled. Liability limits go unexamined because nobody re-reads a declarations page, so a number chosen years ago quietly stops matching the assets sitting behind it. Long-term care is unpleasant to think about, so it goes unplanned.
Coverage gets over-bought when the loss is easy to picture and small: extended warranties on appliances, credit life and mortgage protection policies, disease-specific plans, low deductibles on everything. Each one insures a loss you could absorb, at a price that includes the seller's commission. The cost is not only the premium. It is that the money spent there is money not spent on the coverage that would have mattered.
Which brings up the part of a risk plan that is not insurance at all. Keeping the loss yourself, deliberately and with money set aside for it, is called self-insurance, and it is a legitimate strategy rather than an absence of one. It has two pieces: an emergency fund for the genuinely unexpected, and a sinking fund for the expected but lumpy, like a roof or a transmission. Those two accounts are what make a high deductible safe to choose, which is why they belong in a chapter about insurance. Building them is also the first call on spare cash before any optional policy, a sequencing question our guide to investing takes up.
How does health insurance work, and where do you get it?
A health plan pays in a sequence, and understanding the sequence is most of understanding the plan. You pay the plan's negotiated price yourself until the deductible is met. After that you share costs through coinsurance or copays. Once your own spending reaches the out-of-pocket maximum, the plan pays 100% of covered in-network care for the rest of the plan year.
A hypothetical, to show the sequence
Suppose a plan has a $3,000 deductible, 20% coinsurance, and a $7,000 out-of-pocket maximum, and you need surgery at an in-network hospital where the plan's negotiated price comes to $60,000. You pay the first $3,000. Then you pay 20% of what follows, which reaches $4,000 after another $20,000 of care. At that point you have paid $7,000, you are at the ceiling, and the plan pays the entire remaining balance. Your total for the year is $7,000 no matter how much further the covered in-network bills run: whether the negotiated price was $60,000 or $600,000, your share stopped at the same number. Figures are illustrative only.
Two things never count toward either number, and both catch people out. Premiums do not: they buy the policy and nothing else. Neither, generally, does the balance a provider bills you for going outside the network. So a plan's true worst case in a bad year is twelve months of premiums plus the out-of-pocket maximum, and that total, rather than the premium alone, is what makes two plans comparable.
The network is the biggest single driver of what you actually pay, which is why the cheapest premium is often not the cheapest plan. An HMO generally requires you to stay inside a defined set of providers except in an emergency, often to live or work in its service area, and usually to route specialist care through a primary doctor; a PPO charges more but pays something toward out-of-network care without a referral; an EPO is network-only like an HMO but generally lets you go straight to a specialist. Whatever the letters, the first question about any plan is whether your doctors and your hospital are in it. Plans are also sold in metal tiers, from bronze to platinum, and the tier describes only how the plan splits costs on average. It says nothing about the quality of the care.
Where coverage comes from
There are five routes, and which one is open to you turns on your job, your income, your age, and the calendar.
An employer plan is the default for most working-age households and usually the cheapest, because the employer pays a large share of the premium. That subsidy is also why a spouse's plan is always worth pricing rather than assuming.
The ACA marketplace sells individual plans with a premium tax credit on a sliding income scale, which you can take in advance against the monthly premium and then reconcile against your actual income at tax time. Two current facts change the advice here materially. The temporary enhanced credits expired after 2025, so the credit now ends abruptly above 400% of the federal poverty level rather than tapering, and there is no credit at all above that line. And the caps that used to limit how much a household below the line had to pay back after under-estimating its income were repealed for the same year, so any advance credit you were not entitled to is now repaid in full at every income level. That makes marketplace coverage the one place in personal finance where a single extra dollar of income can cost thousands, and it is why an early retiree considering a Roth conversion or a realized gain should check it against their coverage first. Our guide to taxes covers the reconciliation side.
Medicaid covers people with low incomes. It is run by each state within federal rules, so the income limits, the covered services, and even the program's name vary; states that expanded it cover adults up to 138% of the federal poverty level. Separately, and importantly for a later section, Medicaid is the largest payer of long-term care in the country. Work requirements for some adults take effect under the 2025 federal law around the start of 2027, earlier in states that choose to move sooner and later in states granted extra time, so the rules where you live are worth checking directly.
COBRA lets you keep a former employer's plan after leaving, generally for 18 months, and up to 36 months for events like divorce or a child aging off the plan. You have 60 days to elect it. The reason the price shocks people is that nothing about the plan changed except who pays: you now cover the employer's share as well as your own, plus an administrative fee, up to 102% of the full cost. Because losing job-based coverage also opens a 60-day special enrollment period on the marketplace, the two are worth pricing side by side rather than defaulting to the familiar one.
Medicare becomes the main coverage for most people at 65. Whether you have to do anything about it depends on your situation: if you are already drawing Social Security you are enrolled automatically and mailed a card, and if you are not, you have to sign yourself up, which is where the deadline below bites. It is also not free and not complete. It comes in parts: hospital coverage, which is premium-free for most people with enough work history; outpatient and physician coverage, which carries a monthly premium that rises with income; and prescription drug coverage through private plans. You either add a Medigap supplement to fill the cost-sharing gaps, or take a Medicare Advantage plan, a private alternative that replaces how you receive the main benefits and usually bundles drug coverage. Enrollment runs in a seven-month window around your 65th birthday, and missing it without other qualifying coverage carries a late-enrollment penalty on the outpatient premium of 10% for every 12 months you delayed, for life. Medicare also does not cover long-term custodial care. The full treatment, including the supplement decision and the income-related premium surcharge, belongs to the Government Benefits guide.
High-deductible plans, HSAs, and FSAs
A high-deductible health plan trades a lower premium for more exposure to the first dollars of care, and it unlocks something most plans do not: a health savings account. An HSA is the only account that offers a deduction going in, tax-free growth, and tax-free withdrawals for qualified medical costs. That combination makes choosing a qualifying plan a genuinely different calculation from choosing any other high-deductible plan, because part of what you save on premium can be captured permanently rather than simply spent. The money is yours, it rolls over indefinitely, and it can be invested.
A flexible spending account is the near-opposite and is often confused with it: the plan is the employer's, the money is forfeited at year end apart from a limited carryover or a short grace period (a plan may offer one or the other, never both), and it works with any plan rather than requiring a high deductible. Variants exist for dependent care and for dental and vision only. Smaller employers may instead offer a reimbursement arrangement, giving you a defined amount to buy your own coverage rather than a group plan.
All of it is governed by the calendar. Coverage is bought in windows: an employer's own annual window, usually in the autumn but set by the employer rather than by law, and the marketplace's window, which opens on November 1 each year and is short. The closing date differs between the federal site and state-run exchanges, and a 2025 rule that would have shortened it was struck down in court, so check the current deadline rather than assuming last year's. Outside those windows you need a qualifying life event, such as marriage, a birth, a divorce, a move, or the loss of other coverage, and the deadline to act after one is short. Missing a window can mean a year with no route in.
The supplemental products
Dental and vision coverage are usually separate policies rather than part of a health plan, and they behave differently from one: annual maximums are typically low enough that they function closer to a payment plan for routine care than to protection against catastrophe. That is not a reason to skip them, but it is a reason to compare the premium against the care you actually expect rather than treating them as insurance against a bad year.
A second group is more often misunderstood. Critical illness, hospital indemnity, accident, and accidental death and dismemberment policies pay a fixed cash amount when a specific trigger occurs: a listed diagnosis, a night in hospital, a particular kind of injury. They do not pay for your care and they do not have a network or an out-of-pocket maximum. They can be reasonable supplements next to a real health plan, and they are never a substitute for one. Buying several of them instead of improving the underlying plan is the over-buying error described earlier, because each covers a narrow slice of a risk the main policy already covers broadly.
A few other things a reader will meet, each worth recognising: short-term health plans are not the same as marketplace catastrophic plans, and the difference matters: a catastrophic plan is real marketplace coverage with a very high deductible and an out-of-pocket ceiling, while a short-term plan sits outside most of those protections and can turn you down or exclude a condition you already have. Health care sharing ministries are not insurance and are not regulated as insurance, so nothing in them is a legal obligation to pay. Telehealth and direct primary care arrangements are ways of buying access to a clinician, not coverage against a large bill. And a domestic health plan generally travels poorly, which is why anyone spending real time abroad prices international coverage separately.
What happens if you can't work?
Your income stops and your expenses do not, which is why disability insurance protects the asset that funds every other plan on this page. Consider what a working household actually owns. The house is insured. The cars are insured. The retirement accounts are the visible savings. The largest economic asset, though, is usually the income still to be earned, and a serious illness or injury is what destroys it. That is the asset most often left uninsured.
Three shapes of coverage
Short-term disability covers weeks to several months and is frequently what an employer provides by default. It handles a surgery and a recovery. It does not handle a career-altering diagnosis.
Long-term disability is the coverage that matters, paying for years or through to retirement age. If a household buys one kind of disability insurance, this is the one.
Group coverage is either kind bought by an employer for its workforce. It is inexpensive, usually requires no medical exam, and is the right place to start. It is also narrower than it looks, in five specific ways worth checking on your own policy: it replaces a limited share of pay rather than all of it; it caps the monthly benefit at a figure higher earners exceed; it frequently counts base salary only and excludes bonus or commission income; it may apply a stricter definition of disability after an initial period; and it ends when the job does, which is precisely the moment a health change can make individual coverage expensive or unavailable.
Then there is the tax rule that changes the arithmetic and surprises nearly everyone. If the employer paid the premium, the benefit is taxable income. If you paid it yourself with after-tax dollars, the benefit is tax-free. A group policy advertised as replacing 60% of salary is therefore taxed on the way out, so for a higher earner it can net closer to 40% of gross pay, while an individual policy replacing less on paper may deliver more in the bank. Where employer and employee split the premium, the benefit is taxed in proportion.
How any policy is shaped
Four features decide what a policy is worth. It replaces a percentage of income rather than the whole thing. Part of that is design, since a benefit set below your former income preserves the incentive to go back to work, and part is arithmetic: a benefit you paid for with after-tax dollars arrives tax-free, so a percentage of gross pay lands much closer to your old take-home than the headline suggests. It begins only after an elimination period, a waiting stretch of weeks or months that your emergency fund has to bridge. It pays for a defined benefit period, which may be a few years or run to retirement age, and a longer one costs materially more. And a residual benefit, where offered, pays partially if you return at reduced capacity rather than treating recovery as all-or-nothing.
The single most consequential term in the contract, though, is the definition of disability itself. A policy written on an own-occupation basis pays when you cannot do the job you were trained for. One written on an any-occupation basis pays only when you cannot do any suitable work at all. A surgeon who can no longer operate but could work at a desk is covered under the first definition and not the second, which is why two policies quoting the same monthly benefit can differ enormously in whether they ever pay.
The public backstops, and what they are not
Two government programs are frequently mistaken for coverage. Social Security Disability Insurance exists and pays real benefits, but its standard is strict: you must be unable to do substantially gainful work, and the impairment must be expected to last at least a year or to result in death. Benefits begin only after a five-month waiting period, and the amount is calculated from your earnings record rather than designed to replace a professional income. Workers' compensation pays for injury or illness that arose from the job, which means the test is whether the injury or illness arose out of and in the course of the job, not whether you were on the clock. Most illnesses, and anything unrelated to the work, fall to a private policy. Neither is a plan; both are floors.
Who needs life insurance, how much, and which kind?
You need life insurance if someone would be financially worse off if you died. If nobody would be, you probably do not need it, whatever a sales conversation suggests. That test sounds obvious and it settles most cases: a parent of dependent children, a spouse who relies on your income, a co-borrower on a mortgage, and a business partner all pass it. A single person with no dependents and a retiree whose assets already support the surviving spouse generally do not. The case people most often miss is the stay-at-home parent, whose unpaid work would have to be purchased if they died, which is a real cost even though no paycheck stops.
How much
Work out what the money has to do, then subtract what already exists. On the obligation side: the years of income that would need replacing, the mortgage and other debts, education you intended to fund, and final expenses. On the resource side: your savings, any group life insurance through work and any supplemental coverage you bought through it, and Social Security survivor benefits, which most people forget entirely and which can be substantial for a family with young children. What remains is the gap a policy is being bought to fill. A multiple-of-income rule of thumb is a useful check on that figure and a poor substitute for it, since it knows nothing about how many dependents you have or how much mortgage is left.
Term coverage
Term life insurance is pure death benefit for a set period at a level premium: if you die during the term, your beneficiaries are paid, and if you outlive it the coverage simply ends. That structure fits a need with an expiry date, which is what most needs are, because mortgages get paid off and children grow up. Its variants mostly matter at the edges. Level term keeps the death benefit flat, decreasing term steps it down over time, convertible term preserves your right to switch to permanent coverage later without a new medical exam, and return-of-premium versions refund premiums if you outlive the term, and cost more for that feature.
Permanent coverage, and its family of variants
Permanent insurance does not expire and accumulates a cash value you can borrow against through a policy loan, in exchange for a substantially higher premium for the same death benefit. Within that category the differences are real and are worth knowing before anyone shows you an illustration.
- Whole life fixes the premium and the guaranteed growth of the cash value. It is the most predictable and the most expensive per dollar of death benefit.
- Universal life makes the premium and death benefit adjustable, with the cash value absorbing the internal cost of insurance. That flexibility has a sharp edge: a policy funded lightly in early years can require far larger premiums later to stay in force, which is how policyholders end up with an unexpected bill decades in.
- Indexed universal life ties the cash value's growth to a market index, subject to caps and other limits that mean you do not receive the index's full return.
- Variable universal life puts the cash value into investment subaccounts you choose, which moves the market risk onto you and makes any projection considerably less predictive.
- Survivorship coverage, also called second-to-die, insures two people and pays only after both have died. Because the insurer pays later, the premium is lower than comparable coverage on either person alone, which is why it is used mainly to cover estate costs rather than to replace income.
A handful of narrow-purpose products are worth naming so you recognise them. Final expense policies carry small face amounts and are often issued without a medical exam, priced accordingly. Credit life and mortgage protection policies insure a specific debt rather than your family, with the lender as the practical beneficiary, which is a much narrower promise than a policy your beneficiaries can spend however the need requires. Juvenile policies insure a child, who by definition has no income for anyone to lose.
One fact about distribution belongs here, because a shopper needs it. Permanent policies pay commissions calculated as a percentage of premium, and a permanent premium is several times a term premium for the same death benefit, so the same rate produces a far larger payment. That is why permanent coverage is presented more often and more energetically. That does not make them wrong, and there are genuine uses: an estate holding illiquid assets and facing a tax bill, where a policy owned by an irrevocable life insurance trust supplies the cash; a dependent who will need support for life; funding a buy-sell agreement between business partners. But the recommendation and the compensation move together, and the decades of projected cash value in an illustration are a projection rather than a promise. Where the question is really about savings rather than protection, the honest comparison is against funding a retirement account, which our guide to investing covers.
An existing policy is not a trap. A 1035 exchange moves cash value into another life, annuity, or long-term care contract without triggering tax, though the permitted directions run one way only, so an annuity cannot be exchanged into life insurance; the cash surrender value is what you would actually collect on exit, and over-funding a policy can turn it into a modified endowment contract, which loses part of the tax treatment people bought it for. Any of those is a reason to get the policy analysed rather than to keep paying out of inertia or to cancel out of frustration.
The paperwork that decides everything
A life insurance policy pays whoever is named on the beneficiary designation form, and that form beats the will. A stale designation is one of the most expensive pieces of unfinished paperwork in personal finance: an ex-spouse named twenty years ago collects, and no amount of contrary instruction in a will changes it. Name contingent beneficiaries too, so the money has somewhere to go if the primary beneficiary dies first.
On tax: a death benefit is generally received free of income tax by the beneficiary. It can still be counted in the taxable estate if the insured held any of what the tax code calls the incidents of ownership, and that reaches further than whose name is on it: the right to change the beneficiary, to borrow against the policy, or to surrender it is enough. That is why ownership becomes a live question for larger estates and why the trust arrangement mentioned above exists. Our estate planning guide covers the estate side.
Who pays for long-term care?
Mostly you do, until you have spent down enough to qualify for Medicaid. That is the default outcome in the United States, and the belief that Medicare handles it is the most expensive misunderstanding in retirement planning.
The reason sits in what long-term care actually is. It means help with the activities of daily living, which are eating, bathing, dressing, toileting, transferring in and out of a bed or chair, and continence, provided at home, in assisted living, or in a nursing home. It is assistance rather than treatment, which is exactly why health insurance and Medicare do not pay for it. Medicare covers medical care, including a limited stretch of skilled nursing after a qualifying hospital stay, and nothing resembling years of daily help.
So in practice the bill is paid out of pocket, and then by Medicaid once assets are low enough to qualify. Medicaid is the largest payer of long-term care in the country. Reaching it generally requires spending down assets, and it applies a five-year look-back to transfers made before applying. Afterwards, states are required to seek recovery from the estates of recipients who were 55 or older, at minimum for the long-term-care services Medicaid paid for, and some states go wider. Recovery is deferred while a surviving spouse is alive, and while a child who is under 21, blind, or disabled survives, but deferred is not the same as waived, and whether the claim can still be collected afterwards depends on how your state defines the estate. This is a plan. It is simply a plan most households arrive at by default rather than by choice.
The three deliberate strategies
Self-fund. With enough assets this is the honest answer, and for a meaningful share of households it is the right one. What it requires is that the number be looked at rather than assumed.
Insure. A traditional long-term care insurance policy pays a daily or monthly benefit once you need help. For a policy to qualify for favourable tax treatment, federal law defines the trigger: a licensed practitioner must certify that you cannot perform at least two of the six activities of daily living without substantial assistance for a period expected to last at least 90 days, or that you need substantial supervision because of severe cognitive impairment. That is a meaningful threshold, and it is worth reading before assuming a policy pays when care merely becomes inconvenient to provide.
Hybrid. A life insurance policy or annuity with a long-term care rider trades a smaller care benefit for the certainty that the money is not forfeited if care is never needed. That single objection, paying premiums for decades and possibly getting nothing, is what collapsed traditional sales, and hybrids exist because of it.
Anyone shopping traditional coverage should know three things about the market. Premiums are generally not guaranteed level, and the industry has raised them substantially on policies already in force, so the quoted premium is a starting point rather than a fixed cost. Underwriting looks at your current and past health, and conditions that could lead to care can make a policy unavailable at any price, so waiting costs more than money: it can cost access. And a policy's real value turns as much on its elimination period and its inflation protection as on the headline daily benefit, because a benefit fixed in today's dollars buys much less by the time it is claimed.
The planning point generalises past this one risk. A moderate probability attached to a very large tail is the hardest combination to insure well, which is why this decision resists simple answers. The useful question is not whether you will need care. It is what several years of it would do to the household, including to the spouse who stays, which makes this a decision to work through alongside the rest of a retirement plan rather than separately from it.
What auto coverage do you actually need?
An auto policy is not one product but several coverages sold together, and they protect different people. That is also why "full coverage" is shorthand rather than a product. It is not a coverage part on your policy; it usually means liability plus collision plus comprehensive, the combination a lender requires while it holds a lien on the car. It says nothing about how high your limits are, and it does not include uninsured-motorist coverage, personal injury protection, or gap.
Liability coverage pays for harm you cause other people, split into bodily injury and property damage, and it is where the serious money sits. State minimum limits are a legal floor rather than a measure of adequacy; many were set decades ago and are lower than the cost of a single hospital stay. An at-fault accident that exceeds your limit does not stop there. The remainder reaches your savings and, in many states, your future wages.
Collision and comprehensive cover your own vehicle: collision for a crash, comprehensive for most of the rest, including theft, weather, and hitting a deer. Both are worth carrying while the car is worth enough to matter, and both are worth dropping once the annual premium starts to look large next to what the insurer would ever pay out. That is a calculation about your specific car rather than a rule about mileage or age.
Uninsured and underinsured motorist coverage is the one most drivers do not know they need. It pays when the driver who hit you has no insurance or not enough of it, and between those two that describes roughly a third of the drivers on the road: about one in seven carries no insurance at all, and a further one in six carries too little. Some states require this coverage; many only require it to be offered. Without it, being hit by someone with nothing means your own injuries are your own problem.
Personal injury protection pays medical costs, and sometimes lost wages, for you and your passengers regardless of who caused the accident. It is the primary route to recovery in no-fault states, and it is required in a number of ordinary at-fault states too, so do not assume it is irrelevant where you live. Where it is not required, a smaller version is usually sold as medical payments coverage. How much either matters depends on how good your health insurance already is.
Gap insurance covers the difference between what you owe on a loan or lease and what the car was worth when it was totalled. It matters on a long loan with a small down payment, where the balance can exceed the vehicle's value for years, and it becomes pointless the moment you have equity in the car. Buying it after that point is paying for a risk that no longer exists.
One more thing is worth knowing about this market specifically: prices for identical coverage differ enormously between insurers, and the difference is driven partly by rating factors you influence. Re-shopping the same limits every few years is one of the very few savings available in insurance that costs you no protection at all. Usage-based or telematics programs, which track how you actually drive and price accordingly, are worth a look for low-mileage drivers and worth declining if you would rather not be measured.
What home coverage do you actually need?
A homeowners policy bundles four things: the dwelling itself, the contents, your personal liability, and the extra living costs if the home becomes unlivable. Two choices inside it decide what you actually collect after a loss, and both are easy to get wrong at purchase and expensive to discover later.
The first is replacement cost versus actual cash value. Replacement cost pays to replace the property at today's prices. Actual cash value subtracts depreciation, which is why a twelve-year-old roof pays out a fraction of what a new one costs. On a replacement-cost policy some of the payment is often held back until the work is actually done, which catches homeowners who expected a single cheque.
The second is how much of the replacement cost you insure, and this one has a mechanism people rarely hear about until they claim. Most policies require you to insure the dwelling for at least 80% of its full replacement cost in order to be paid replacement cost on a loss. Fall below that and the insurer pays proportionally, on partial losses as well as total ones. So under-insuring does not simply cap what you would get in a fire that destroys everything. It quietly reduces what you collect on the kitchen you actually had to repair.
Then there are the coverages that are not gaps in your policy so much as separate markets. Flood is excluded from every standard homeowners policy and bought as its own policy, priced on the building itself rather than on its flood zone. Earthquake is likewise separate, and typically carries a deductible expressed as a percentage of the coverage limit rather than a flat sum, which makes it much larger than homeowners deductibles people are used to. Every year, households discover both of these after the event rather than before it.
Renters insurance exists because a landlord's policy covers the building and nothing of yours. A renters policy covers your possessions, your personal liability, and your living costs if the unit becomes uninhabitable, for one of the lowest premiums in insurance. Landlord coverage is the mirror image: a homeowners policy stops applying once you rent the property out, and the landlord form adds loss of rental income while excluding the tenant's belongings.
Title insurance is the policy most buyers pay for without ever being told what it does. It covers defects in the property's ownership history: a forged signature, an unpaid contractor's lien, unpaid taxes, an heir nobody knew about. Unlike everything else on this page it insures the past rather than the future, and it is paid with a one-time premium at closing rather than annually. The critical distinction is that the lender's policy your mortgage requires protects the lender's loan, not your equity. A separate owner's policy is the one that protects you, and it is optional, which is why many buyers pay for title insurance at closing and walk away with none of their own.
Private mortgage insurance is included here because it is routinely mistaken for coverage that protects the borrower. It protects the lender against your default, and you pay for it. The upside is that it ends: on a primary residence you may request cancellation once the balance reaches 80% of the home's original value, and it terminates automatically at 78% if you are current on payments. FHA mortgage insurance follows different rules and frequently cannot be removed without refinancing, which is a genuine cost to weigh when comparing loan types.
Finally, a home warranty is not insurance at all. It is a service contract covering repairs to appliances and systems, which places it squarely in the category to handle with cash rather than a policy: the losses are absorbable, and the price reflects the seller's margin rather than a catastrophe.
Why is liability the coverage people get most wrong?
Because property coverage caps at the value of the thing, while liability has no natural ceiling. A destroyed car costs you a car. A serious accident in which you are at fault can cost more than the car, the house, and the retirement accounts together, because the bill is someone else's medical care, rehabilitation, and lost earnings rather than the price of anything you own.
There is a structural problem underneath that. Personal liability coverage is not bought as one thing. It arrives in pieces, attached to the auto policy and the homeowners or renters policy, each with its own limit set at a different time for a different reason. What matters is not the highest of those limits but the lowest, because a claim arrives through whichever policy happens to apply. A household can carry a respectable limit on the house and a decade-old default on the cars, and it is the cars that will be driven.
An umbrella policy is the fix. It sits on top of the underlying policies and adds a layer of liability coverage above their limits, typically a million dollars or more. Insurers require you to raise those underlying limits to a minimum level first, so buying one usually improves the base policies as a side effect. The premium is generally modest relative to the coverage, precisely because claims that large are uncommon, which makes an umbrella the highest coverage-per-dollar purchase available to most households. It also reaches some claims the base policies exclude, such as certain personal-injury actions like libel or slander.
The test for whether you need one is straightforward: add up what you have plus what you will earn, and compare it against your current liability limits. Most people cross that line without noticing, because assets accumulate quietly and limits do not move on their own. One related term is worth recognising if you own a business and are offered something similar: excess liability coverage follows the underlying policy's terms rather than broadening them, so it raises the ceiling without adding the extra reach an umbrella gives.
The exposures households forget are consistent enough to list: a teenage driver, a rental property, a dog, a swimming pool or trampoline, a boat, hosting a party where alcohol is served, serving on a nonprofit board, and what a family member posts online. None of these changes the property side of your insurance. All of them change the liability side.
What if you work for yourself or own a business?
Going self-employed does two things at once. It removes coverages an employer was quietly providing, and it creates exposures that personal policies expressly exclude. The second half is the one that surprises people: a homeowners policy covers very little done for money. Business property in the home is typically capped at a token sublimit, and business-related liability claims are generally excluded outright, so a client who trips on your step is not a homeowners claim.
Two coverages answer genuinely different losses and are the pair most often confused.
General liability, written formally as commercial general liability, covers bodily injury and property damage your operations cause to somebody else, along with personal-injury claims like libel and slander. This is the coverage a client's contract or a commercial landlord's lease will name by name, and the one that adds the client to your policy as an additional insured when they ask.
Errors and omissions, also sold as professional liability and as malpractice coverage in some fields, covers the different claim that your work caused a client financial harm: bad advice, a design that failed, a filing error, a missed deadline. General liability does not touch that claim, because nobody was physically hurt and no property was damaged. That is why anyone selling advice or professional services generally needs both rather than choosing between them, and why discovering the distinction after a claim is a bad way to learn it.
Workers' compensation pays an injured employee's medical costs, rehabilitation, and lost wages, and in exchange limits their right to sue you. Nearly every state requires it, though not all do and the shape of the requirement varies: Texas lets most private employers opt out, South Dakota has no requirement at all, and Wyoming's mandate reaches enumerated hazardous occupations with coverage elective for the rest. Beyond that, the details, including which employers are covered and how small an operation has to be to fall outside the requirement, are set by state law rather than federally, so this is one to confirm with your own state rather than from any national summary. Separate federal programs exist, but they cover specific groups such as federal employees, longshore and harbor workers, and coal miners rather than private employers generally.
Then the coverages that follow from having a business at all, briefly each. Commercial property covers the building, equipment, inventory, and furnishings your personal policy excludes. Business interruption replaces the earnings lost while you cannot operate after a covered physical loss such as a fire, and pays continuing expenses like payroll and utilities: it covers the income, not the damaged property, which is a distinction worth holding because the two are bought together and claimed separately. Note that these policies generally start paying after a short waiting period rather than from the moment of the loss. Commercial auto exists because a personal auto policy will not follow you into every kind of business driving. Most personal policies exclude driving for hire, though insurers now sell rideshare endorsements that put some of it back, and paid delivery work usually runs into the same for-hire language. A vehicle titled to the business needs the business named as the insured, so a personal policy in your own name is where a claim on a company car gets disputed. Employment practices liability covers claims by employees alleging wrongful termination, discrimination, harassment, or failure to promote, none of which a general liability policy covers. Cyber liability covers what follows a data breach or ransomware attack for a business holding customer information, which standard policies generally do not.
Two more exist because a small business often is a person. Key person insurance is a life insurance policy the business owns on someone it depends on, paying the business rather than a family if that person dies, so it can absorb the disruption or fund a transition. Business overhead expense insurance pays the business's fixed costs, the rent and the utilities and the staff wages, while the owner is disabled. It is a different policy from the personal disability coverage that replaces the owner's own income, and an owner who needs both and buys one has covered half the problem.
Most small operations do not assemble all of this separately. A business owner's policy packages the common coverages, typically property, business interruption, and general liability, often for less than buying them individually, and it is the usual starting point. Two things are worth checking before treating it as done. The package usually sets one combined liability limit, which is simpler to choose but less flexible, and it may not be enough for a higher-risk operation: higher limits or an umbrella policy are the fix. And note what such a package generally leaves out, because the omissions are the ones that get forgotten: commercial auto, workers' compensation, professional liability, and health or disability coverage.
One scoping point, since it comes up constantly: forming an LLC is not a substitute for insurance. An entity can limit certain liabilities, and it does not stop a claim from reaching the business's own assets, nor shield a professional from a claim about their own work. Entity choice and insurance solve different problems, and a business needs an answer to both.
Which of the smaller coverages are worth it?
Apply the same severity-and-absorbability test, and the answer becomes a method rather than a list of verdicts. Ask what the worst case actually costs, whether you could absorb it, and how much of the premium is buying protection against a real catastrophe as opposed to the convenience of not writing a cheque. Most of the smaller policies fail the second question, which is what makes them poor value. A few pass it for particular people.
Travel insurance earns its place when the prepaid, non-refundable cost of a trip is large enough to hurt, or when you are going somewhere your health plan does not follow. The piece most worth prioritising is usually the medical and evacuation coverage rather than the trip-cancellation coverage: the State Department puts an air-ambulance evacuation back to the United States at anywhere from $20,000 to $200,000 depending on where you are, and notes that most plans will not pay for one. A cancelled trip, by contrast, costs you only what you prepaid. If that prepaid figure is itself large, a cruise or an escorted tour, cancellation coverage earns its keep too, and it is the piece far more likely to be claimed.
Pet insurance is a real answer to a real problem: the four-figure emergency veterinary bill that otherwise goes on a credit card, or forces a decision nobody wants to make on the day. Two things are worth weighing against each other. For a healthy young animal, several years of premiums can exceed the four-figure bill you were insuring against, which is the argument for funding it yourself in a sinking fund instead. But what insurance is really buying is the rarer five-figure bill, where no plausible run of premiums and no realistic sinking fund competes. Either way you never recover the whole invoice, because deductibles and coinsurance apply, and pre-existing conditions are generally excluded.
Identity theft insurance reimburses the cost of cleaning up the mess, meaning postage, notarisation, lost wages, and legal fees, rather than the money that was stolen. Policy limits typically run in the low five figures, and the coverage frequently overlaps with a homeowners or renters policy you already hold, so check that before buying it separately. Your direct exposure is limited to begin with: federal law caps your liability for unauthorised credit-card charges at $50, and many card agreements waive even that. Debit cards work differently, and how much you can be left with depends on how quickly you report the loss, which is the real argument for reading statements promptly. Meanwhile the most effective preventive step is free. Federal law entitles you to place and lift a credit freeze at each of the three nationwide credit bureaus at no charge, and it stops a new credit account being opened in your name. Two limits on that: a freeze does not protect accounts you already have, and it does not stop prescreened credit offers, which need their own opt-out.
Extended warranties, home warranties, tuition insurance, and wedding insurance share a shape: they cover losses most households could absorb. The Federal Trade Commission's own guidance on extended warranties and service contracts makes the case plainly, noting that such a contract may not be worth the cost if the product is unlikely to need repairs, that it is poor value if it adds nothing to the warranty that came with the product, and that putting the money into a savings account instead may be the better option. That is the self-insurance argument, from a federal regulator. If a broken appliance would be an annoyance rather than a crisis, the money is better kept. Note also that a home warranty is generally not regulated as insurance at all, which varies by state and changes what recourse you have if a claim is refused.
The protection you already have and did not buy
Three safety nets exist without a purchase, and readers consistently assume they are broader than they are.
FDIC and NCUA coverage protects deposits at insured banks and credit unions, currently $250,000 per depositor, per institution, per ownership category, which is why the same person can be covered for more than that across joint and individual accounts. What neither covers is what you bought through the institution: not stocks, bonds, or mutual funds, not annuities, not life insurance, and not the contents of a safe deposit box. The FDIC's published list also names crypto assets explicitly.
SIPC protects the custody function at a failing brokerage. Its $500,000 per customer, including a $250,000 limit for cash, is not a pot of coverage paid out on top of what you owned: customer property is gathered and distributed first, and the figure caps an advance that fills whatever shortfall remains. Two things about it are widely misunderstood. It is not a government agency, and in its own words it "is not the securities world equivalent of the Federal Deposit Insurance Corporation"; and it does not protect the value of any security. If your broker collapses, SIPC works to return the securities and cash that were in your account. If your fund falls 30%, nothing about SIPC applies.
State guaranty associations back insurance policies up to state-set limits if the insurer itself becomes insolvent. Those limits are real, and they are finite, which is why an insurer's financial-strength rating is worth checking on any promise you are relying on decades out. The reason it matters more there than on a car policy is not the length of the contract but the lock-in: with long-term care coverage or a lifetime income contract you pre-pay for a benefit you may not claim for thirty years, you cannot switch without forfeiting what you have accrued or being re-underwritten at an older age, and a large contract can exceed the guaranty limit. With auto insurance you change carriers at renewal for nothing.
Common insurance mistakes
These recur across households of every income level, and each one has a specific cost rather than being merely untidy.
- Insuring the small things while leaving the large ones bare. An extended warranty on a dishwasher and a low auto deductible, alongside state-minimum liability limits, is a household that has spent real money buying comfort and left the catastrophe uncovered.
- Never re-reading the declarations page. Coverage limits are set once and then age. A dwelling figure from when you bought the house, a liability limit from a decade ago, and no umbrella after your net worth grew are all the same mistake: the policy is now sized for a life you no longer have.
- Treating the disability policy at work as the answer. Without knowing the replacement percentage, the monthly cap, whether bonus income counts, the definition of disability after the first couple of years, and whether the benefit will be taxed, you do not actually know what you have.
- Comparing health plans on premium alone. The comparable figure is a year of premiums plus the out-of-pocket maximum, checked against whether your doctors are in network. Plans frequently reverse rank once you do that.
- Assuming Medicare covers long-term care. It does not, and the assumption is usually discovered during a crisis, when the options are worst and the time to plan has gone.
- Riding COBRA past the Medicare enrollment window. This buys a permanent late-enrollment penalty on Medicare's outpatient premium, for a reason nobody would guess: COBRA does not count as coverage based on current employment.
- Assuming the enhanced marketplace subsidies still apply. They expired after 2025. A modest increase in income above the cutoff now eliminates the credit outright rather than reducing it. Separately, the caps that limited repayment for households below the cutoff were repealed for the same year, so an income estimate that turns out low is now repaid in full whatever you earn.
- Buying permanent life insurance as an investment without pricing the term alternative. The comparison is not difficult to run, and a projected cash-value column is an illustration rather than a guarantee.
- Leaving a stale beneficiary designation. The form beats the will. An ex-spouse named years ago and never updated collects the death benefit, and no contrary instruction elsewhere changes that.
- Letting a policy lapse to save a premium. Reinstating is not symmetrical with cancelling. If your health changed in the interval, replacement coverage may cost far more or be unavailable at any price.
- Assuming flood is covered by homeowners insurance. It is excluded from every standard policy, as is earthquake.
- Mistaking private mortgage insurance for your own protection. You pay it, and it protects the lender. Knowing that is what prompts people to ask about cancelling it once they have the equity.
- Going self-employed and assuming a homeowners policy or an LLC covers business work. Neither does. Personal policies exclude business liability, and an entity is not insurance.
- Carrying general liability while doing professional work. It leaves the claim most likely to arrive, that your advice or your work product cost a client money, entirely uncovered.
When is professional help worth paying for?
When the amount of coverage is a genuine question rather than a form to complete. Plenty of insurance decisions do not need anyone: buying renters insurance, adding a car to a policy, or choosing between two employer health plans at open enrollment are jobs you can finish yourself with the total-cost comparison described earlier.
Advice earns its cost in a smaller number of situations, and they are recognisable. Sizing life and disability coverage when a family depends on the income, because the number depends on your obligations, your existing resources, and your tolerance for leaving a gap. Evaluating a permanent life or long-term care proposal you already have in hand, which is the single highest-value use of an hour of independent advice, because the proposal was designed to persuade and reading it properly takes practice. The Medicare transition, where the supplement decision has a timing asymmetry that is expensive to discover several years late. A household whose net worth has quietly outgrown its liability limits. And the newly self-employed, where the gaps are structural rather than a matter of degree.
One structural fact about this corner of financial services belongs here, because it shapes the advice you are likely to receive. Insurance is where a recommendation and a payment are most tightly joined: a policy generally pays a commission at the point of sale, and the largest commissions sit on the most complex products. That does not make a licensed agent's recommendation wrong, and an agent is frequently the only route to a policy at all. It is worth knowing how the person advising you is compensated, and worth knowing that an independent agent or broker represents several insurers while a captive agent represents one. A conflict of interest is not an accusation; it is a structural fact to account for, and the usual way to account for it is a second opinion from someone who is not paid by the sale.
Where to read further, depending on which part of that you want to pin down: what a fiduciary obligation does and does not require, how the advice-only model differs from other ways of paying for advice and how it compares with fee-only, and our guide to finding an advisor, which covers how to check anyone's registration and disciplinary history before you hire them. You can also browse advisors who work on insurance and risk management.
Key terms in insurance
The vocabulary that shows up on policies, statements, and every article on this topic, each defined in plain English in our glossary.
- Term Life Insurance
Term life insurance is pure death-benefit coverage: you pay a level premium for a set period (commonly 10, 20, or 30 years), and if you die during that term, the insurer pays your beneficiaries a tax-free lump sum. Outlive the term and the coverage simply ends.
- Disability Insurance
Disability insurance replaces part of your income if illness or injury keeps you from working. It protects the asset most working people never think to insure: their ability to earn a paycheck for the next few decades.
- Long-Term Care Insurance
Long-term care insurance pays for extended help with daily living (home aides, assisted living, nursing care) that health insurance and Medicare largely do not cover. Policies pay out when you can no longer perform basic activities of daily living or suffer cognitive impairment.
- Health Savings Account (HSA)
A health savings account (HSA) is a tax-advantaged account for people with high-deductible health plans that offers a triple tax break: deductible contributions, tax-free growth, and tax-free withdrawals for qualified medical expenses.
- Annuity
An annuity is a contract with an insurance company: you pay a premium, and the insurer promises a stream of payments, often for life. It is the only private product that insures against outliving your money, but costs, surrender charges, and commission-driven sales make careful evaluation essential.
- Longevity Risk
Longevity risk is the risk of living longer than your money lasts. It is not a risk of markets but of arithmetic (the longer a retirement runs, the less any given portfolio can safely pay each year), and it is the one retirement risk that gets worse the better things go.
- Fixed Indexed Annuity (FIA)
A fixed indexed annuity is an insurance contract that credits interest based on the movement of a market index, subject to caps and other limits, and credits zero rather than a loss when the index falls. It is generally not an SEC-registered security; it is regulated as insurance under state law.
- Survivor Benefits
Survivor benefits are payments that continue to a spouse, child, or other dependent after someone dies. They are not one program but a category — Social Security, employer pensions, the military, annuities, and life insurance each pay them under their own rules, and most of the decisions that determine what a survivor receives are made years before the death.
- Beneficiary Designation
A beneficiary designation is the form on file with your retirement account, life insurance policy, or bank account naming who receives the money at your death. It overrides your will, which is why an outdated form is one of the most common and costly estate planning mistakes.
- Emergency Fund
An emergency fund is cash set aside to cover genuine surprises, a job loss, a medical bill, a failed transmission, so they don't land on a credit card or force you to sell investments at a bad time. The common target is three to six months of essential expenses.
- Sinking Fund
A sinking fund is money set aside a little at a time for a specific, predictable future expense (like insurance premiums, holiday gifts, or car repairs), so the bill arrives already paid for.
Browse all 33 insurance terms in the glossary, or start from the Guide to Personal Finance.