The phrase comes from the insurance trade, long before economics picked it up, and it names something insurers observed constantly: a person with a fully insured warehouse is less careful with it than a person without.

Not because insurance attracts arsonists. Because caution has a cost, the benefit of caution has been transferred to the insurer, and the calculation on the ground has genuinely changed.

The structure is always the same: whoever makes the decision does not carry the full consequence.

Where it shows up in a small business

The unlimited retainer. A fixed monthly fee with undefined scope makes every additional request free at the point of asking. Requests multiply — reasonable ones, from reasonable clients — and margin erodes. The client is responding correctly to the price they face, which is zero. This is most of why pricing a retainer is a scope problem rather than a number problem.

Salary without exposure. An employee proposing an expensive initiative bears none of it if it fails. That is the principal-agent problem and moral hazard is the specific form it takes around risk: the upside is shared and the downside is not.

Someone else's budget. Spending is less careful when the money is the company's, the client's, or the investor's. Every expense policy is a response to this.

Guarantees and refunds. A money-back guarantee moves the risk of a bad outcome onto the seller, which is exactly what makes it a strong offer — and it does change buyer behavior at the margin, which is the cost of making it.

The part worth getting right

Moral hazard is not a reason to refuse to carry risk. Carrying risk for a client is frequently the product: it is what a guarantee sells, what a fixed price sells, and what an insurance policy is.

The useful question is narrower. When you take on someone's downside, what behavior have you just made free for them? Then price that, bound it, or share it.

The three standard answers, all of them partial:

A deductible. Leave some of the downside with the party making the decision. An excess on a policy, a minimum commitment, a shared cost. Small amounts change behavior out of proportion to their size.

A bound. Define the scope. Not to be difficult, but because an undefined commitment is an unpriced one, and unpriced commitments are the ones that end relationships.

Monitoring. Expensive and it works, which is why insurers inspect and why a contract with unusual exposure comes with unusual reporting.

The asymmetry to watch

The case that does real damage is the one where the upside stays and only the downside moves — the trader with a bonus on gains and a salary floor on losses, the manager whose successful bets are promotions and whose failures are absorbed by the company.

That structure reliably produces more risk-taking than anyone intended, and it does so through people behaving sensibly given what they face. When a decision consistently comes out more aggressive than the company would choose, the first place to look is who holds the loss.