Moral Hazard Explained (With Insurance Examples)

Moral hazard is the tendency to take on more risk, or take less care, once someone else bears the cost of a bad outcome. It appears after an agreement is signed, when one side can change their behavior in a way the other side cannot observe. An insured driver who stops locking the car, a bank that makes riskier loans because it expects a rescue, and an employee who coasts because effort is invisible are all the same problem wearing different clothes.

The idea sits at the center of insurance markets, banking regulation, employment contracts, and health policy, which is why it shows up in economics, finance, law, and business courses alike. This guide works through the arithmetic that produces it, separates it from the concept it is most often confused with, and covers the standard fixes.

Why moral hazard happens

Taking care is costly. Locking a bike takes time, servicing a car takes money, screening a loan applicant takes effort. A person takes care when the benefit of care, measured as the reduction in expected loss, exceeds the cost of taking it.

Insurance changes that calculation by moving the loss onto someone else. The cost of care stays exactly where it was, but the benefit of care, from the insured person's point of view, shrinks or disappears. Care that was worth taking becomes care that is not worth taking. Nobody has to be dishonest for this to happen, which is why the word moral is misleading. It is an incentive problem, not a character problem.

The insurance arithmetic

A bike is worth $1,000. Without a lock it is stolen with probability 0.20, so the expected loss is $200. A $60 lock cuts the theft probability to 0.05, bringing expected loss down to $50. The lock therefore reduces expected loss by $150 while costing $60, so buying it creates $90 of value. Whether the owner actually buys it depends entirely on who pays for the theft.

Read the middle row carefully, because it is the entire concept. Under full coverage the owner loses nothing whether the bike is stolen or not, so the lock saves them nothing personally and the $60 is pure cost. They rationally skip it, theft rises from 5 percent to 20 percent, and society loses the $90 of value the lock would have created. The insurer eventually charges everyone a higher premium to cover the extra thefts.

The bottom row shows the standard fix. With a $500 deductible the owner still carries real money into every theft, so the lock saves them 0.20 times $500 minus 0.05 times $500, which is $75. That beats the $60 cost, and the lock goes back on.

Moral hazard vs adverse selection

These two are the most commonly confused pair in the whole topic, and the distinction is simply a matter of timing and of what is hidden.

A useful test: ask whether the person was always like that, or whether the contract changed how they behave. If a driver was already reckless and insurers could not tell, that is adverse selection. If the driver became reckless after buying full coverage, that is moral hazard. Both are consequences of asymmetric information, which is the parent idea.

Where it shows up beyond insurance

Banking. A bank that expects a government rescue if it fails keeps the profits from risky lending while the public absorbs the losses. That asymmetry pushes banks toward more risk than they would otherwise accept, which is why regulators impose capital requirements: forcing shareholders to keep real money at stake restores the deductible.

Employment. An employer cannot watch every hour of effort. If pay is fixed and effort is unobservable, the marginal cost of working harder falls on the employee while the marginal benefit goes to the firm. Commission, bonuses, and equity exist to put some of the outcome back on the person taking the action.

Health care. When a visit costs a patient nothing at the point of use, the patient's private cost of an extra appointment is close to zero even though the appointment consumes real resources. Co-payments are the deductible logic applied to medicine.

Rental and shared property. Renters treat cars and apartments less carefully than owners do, which is why security deposits exist.

In each case the pattern is identical. Someone chooses an action, someone else absorbs the consequence, and the action cannot be observed cheaply enough to write into the contract.

How markets and governments respond

Every real fix works by putting some of the risk back onto the person making the decision, or by making the hidden action less hidden.

Deductibles, co-payments and coinsurance leave a slice of every loss with the insured party. This is why full-coverage insurance is rare: it is not that insurers cannot afford it, it is that the product would destroy the behavior it depends on.

Monitoring attacks the information problem directly. Telematics boxes in cars, software that logs work, and bank supervision all convert a hidden action into an observable one, at a cost.

Performance-linked pay ties the agent's reward to the outcome the principal cares about, which is the standard answer to the principal agent problem.

Reputation and repetition matter when the same parties deal with each other repeatedly. A contractor who wants future work has a reason to take care that a one-time contractor does not.

None of these eliminates moral hazard. They trade it against something else: deductibles reduce the risk-sharing that made insurance valuable, and monitoring costs real resources. The efficient contract balances the two rather than removing the problem.

Common mistakes

Treating it as dishonesty. Moral hazard does not require anyone to lie or cheat. The bike owner who skips the lock under full coverage is responding correctly to the incentives in front of them. Exam answers that describe it as fraud have missed the mechanism.

Getting the timing backwards. If the answer involves who chooses to buy the product, it is adverse selection. If it involves how someone behaves once they hold the product, it is moral hazard. Write down which one the question is asking before reaching for a fix, because the fixes are different.

Assuming the fix is to abolish insurance. Insurance exists because risk-averse people gain from pooling risk, and that gain is real. The presence of moral hazard argues for partial coverage, not for none.

Confusing it with an externality. They are related, since both involve costs landing on someone who did not choose them, but an externality falls on third parties outside the transaction, while moral hazard falls on the counterparty inside it.

Practice and connect

Moral hazard is one of the four standard sources of market failure, alongside externalities, public goods, and market power. Make sure you can state the definition in one sentence, separate it from adverse selection on timing, and name a fix that restores the decision-maker's stake. Reinforce the terms in the moral hazard and asymmetric information glossary entries, then work through the wider topic in the market failure module.