Self-Insurance

Self-insurance is the practice of retaining a risk rather than transferring it to an insurer. In personal financial planning it describes any deliberate choice to absorb a potential loss out of your own resources: carrying a $2,500 deductible instead of a $500 one, declining collision coverage on a car worth little, skipping an extended warranty, or simply accepting that a three-thousand-dollar repair will be paid from savings when it arrives. Insurance is the transfer of a risk for a price, and self-insurance is the decision that the price is not worth paying.

The phrase is worth pinning down, because it names three different things and only one of them is a household decision. This is the planning-profession usage, which describes a household's own risk-management choice rather than a regulatory status. The second meaning is a self-funded or self-insured employer health plan, in which an employer pays claims out of its own funds rather than buying coverage, typically with stop-loss protection behind it and largely outside state insurance regulation because ERISA governs it; the tax regulations use exactly that vocabulary, at 26 CFR 1.105-11, "Self-insured medical reimbursement plan." The third is statutory self-insurance, a regulated status a business qualifies for by demonstrating financial responsibility instead of buying a policy, which comes with net-worth tests, reserves and approvals. Federal contractors, for example, must submit a proposed self-insurance program to a contracting officer and obtain approval when the amounts involved are large enough (48 CFR 28.308), and underground storage tank owners can satisfy their obligations through a financial test (40 CFR 280.95). Neither of the last two is available to, or relevant to, an individual.

The decision rule is about the shape of the loss, not its likelihood. Insurance is worth buying when a loss is severe enough that absorbing it would change your financial life, and rare enough that the premium is a small fraction of the loss. It is a poor purchase when the loss is small, frequent, or so predictable that the premium simply prepays it with the insurer's costs and margin added. That is why liability coverage is worth a great deal of money and an extended warranty on a $400 appliance is not: one has a tail that reaches your assets and future income, and the other has a ceiling you already know.

The corollary is the constraint that does most of the work. A risk can only be retained if the money to absorb it genuinely exists, in a form that can be reached without selling something at the wrong time or borrowing at a bad rate. Nobody self-insures a $600,000 liability judgment, a decade of lost earnings from a disability, or the total loss of a home, because no household balance sheet answers those. The practical test is not whether you are willing to bear the loss but whether you could write the check.

Deductibles are where this shows up most often, and they are measurable. Raising a deductible is a partial transfer of risk back to yourself in exchange for a lower premium, and the trade can be evaluated arithmetically: divide the extra exposure by the annual premium saving and you get the number of claim-free years needed to break even. That calculation only means something if the higher deductible is money you actually have available on the day of the loss, which is the connection to a cash reserve. It is not the same thing as a reserve, though. An emergency fund answers "where is the money"; self-insurance answers "what is standing behind this risk." A household can have a large fund and still be badly under-insured, and a household with no fund is not self-insuring anything, it is exposed.

Two further honest points. First, self-insurance loses the other things a policy buys besides money. A liability policy provides a legal defense as well as indemnity, and a household bearing that risk itself has to hire and pay for its own. Second, retaining a risk is not the only alternative to transferring it. Risk can also be avoided, by not owning the boat, and reduced, by installing the water shutoff valve or the monitored alarm. Reduction is frequently the cheapest of the three and gets considered least, partly because nobody sells it.

A household lists the losses it could face, sorts them by how bad the worst case is and how often the loss occurs, and decides for each whether to buy a policy, buy a policy with a larger retention, or carry the risk outright. The decision is revisited when either the exposure or the household's ability to absorb it changes, which in practice means when an asset depreciates, a balance sheet grows, or a household's income becomes more or less stable.

A hypothetical, on deductibles. Suppose an auto policy costs $1,640 a year with a $500 deductible and $1,460 a year with a $1,000 deductible. The saving is $180 a year and the additional amount retained is $500 per claim. Divide $500 by $180 and the break-even is about 2.8 years: if a claim occurs less often than roughly once every three years, the higher deductible costs less over time. The caveat is not statistical but practical, and it is the whole point of the exercise. The trade is only sound if the extra $500 is available on the day of the accident.

A second hypothetical, on dropping a coverage entirely. Suppose a car has an actual cash value of $2,800 and its collision coverage carries a $1,000 deductible and costs $420 a year. The most that coverage can ever pay is $2,800 minus $1,000, or $1,800, and that is only in a total loss. Paying $420 a year for a maximum recovery of $1,800 is a ratio a household can weigh directly, and it is why collision is commonly dropped on an old vehicle. The arithmetic is entirely different on a liability coverage, where the maximum recovery is not bounded by the value of anything you own.

The pattern in both examples is the same. Where the worst case is a known, bounded number the household can absorb, the arithmetic is a genuine comparison. Where the worst case is open-ended, there is no comparison to make.

Pros

• It removes the insurer's expenses and margin from risks that do not need transferring, which is real money on small, frequent, predictable losses.
• Higher deductibles reduce premiums immediately and measurably, and the trade-off can be calculated rather than guessed at.
• It avoids paying for coverage whose maximum payout is small relative to its premium, which is the usual case on aging property and low-value items.
• Making the decision explicitly forces a household to look at its actual exposures, which is the part that most often turns up a genuine gap.
• No claim, no adjuster, no dispute over whether a loss was covered.

Cons

• The retained loss lands all at once, and it lands at the worst moment roughly as often as at a convenient one.
• It only works if the money is genuinely liquid. A retention notionally covered by an investment account may have to be funded by selling in a bad market or borrowing on a credit card.
• Households systematically underestimate low-probability severe losses, which is exactly the class that should never be retained.
• Self-insuring a liability risk also forgoes the insurer's duty to defend, so legal costs come out of the same pocket as the judgment.
• It requires discipline. The premium saved has to actually stay saved, and frequently does not.