Fundamentals

Moral Hazard: What It Is, How It Arises, and Why Incentives Matter

By Daniel Sardá · Published on October 13, 2026

6 min read1,223 words

Moral hazard arises when someone can change their behavior after receiving protection or funding without bearing the full cost of their decisions.

What changes when someone can make a decision while knowing they will not bear the full cost if things go wrong? That question lies at the center of moral hazard, an incentive problem found in insurance, lending, firms, and bailout policies.

The term can be misleading. It does not necessarily describe immoral conduct or imply bad intent. In economics, it refers to a situation in which one party changes relevant behavior after receiving protection, funding, or authority, while another party bears part of the cost. The problem is harder when that behavior cannot be fully observed or specified in a contract.

What moral hazard means

Moral hazard arises after an agreement. A person, firm, or institution obtains some protection against a loss and may therefore have less incentive to prevent it or more incentive to take risks.

Three questions help identify the mechanism:

• Who makes the decision that affects the risk?
• Who can observe or verify that behavior?
• Who pays if the outcome is unfavorable?

Moral hazard exists when the answers do not fully align: someone makes the decision, their behavior is difficult to observe, and part of the cost falls on someone else. The economic theory of moral hazard examines this sequence of contract, private action, and observable outcome.

Key idea: Moral hazard is not an accusation about a person's character. It is a possibility created by the way information, decisions, and costs are distributed.

Nor is a loss alone enough to establish moral hazard. An accident, a default, or a failed project may result from chance. For the concept to apply, the protection must have been capable of influencing behavior related to the outcome.

Insurance as the basic example

Imagine someone who buys auto insurance. Before obtaining coverage, they bear nearly all the loss from theft or a collision. Afterwards, the insurer covers a substantial part of that cost. This protection is valuable: it transfers a risk that could be burdensome for one person to carry alone.

But coverage can also change the incentive to take precautions. The insured person may be slightly less careful about where to park or how much to invest in prevention. They need not lie, cause an accident, or act recklessly. It is enough that reduced exposure to loss changes their behavior at the margin.

The insurer can observe a claim, but not every daily decision. Because the outcome combines behavior and chance, an insurance contract uses deductibles, coverage limits, premium discounts, or verifiable conditions rather than demanding impossible “perfect prudence.”

The aim is not to eliminate insurance in order to prevent every behavioral change. It is to preserve its protective function without completely disconnecting decisions from their consequences.

From lending to expected bailouts

The same mechanism appears in other relationships, although it takes different forms.

In lending, a creditor provides funds on certain terms, but the borrower may later choose how to use them. If losses fall mainly on the creditor while potential gains benefit the borrower, the borrower may favor a riskier project than the one initially anticipated. Collateral, monitoring, and staged disbursements seek to narrow that gap. Such arrangements also show how economic incentives shape decisions.

Within a firm, owners may delegate decisions to managers whose effort and choices are not fully observable. If compensation rewards some outcomes while excluding other costs, managers may respond to those signals even when their conduct does not serve owners’ long-term interests.

Moral hazard can also occur when banks, firms, governments, or other actors expect to be rescued if they face severe losses. The expectation of support may weaken prior discipline and encourage greater exposure to risk. This does not mean every bailout has that effect or can never be justified. It is an institutional tension: protection during a crisis may be necessary, but a broad and predictable promise of protection can change the decisions that precede it.

Caution: Support, a guarantee, or a subsidy does not automatically create moral hazard. It is necessary to show that the expectation of protection changes behavior and shifts part of its cost.

Moral hazard, adverse selection, and fraud are not the same

These concepts often appear together because all concern incomplete information, but they answer different questions.

Adverse selection occurs before a contract is made. One party knows a relevant characteristic—such as its level of risk—that the other party cannot readily distinguish. The question is: “what kind of risk enters the agreement?”

Moral hazard occurs afterwards. The decisive issue is not only who the person is, but what they do once they have coverage, money, or authority. The question becomes: “how does their behavior change?” The two problems can coexist.

Fraud, by contrast, requires deliberate deception to obtain an illegitimate advantage. Taking fewer precautions because someone else bears part of the cost can be moral hazard without involving falsehood or a deliberately caused loss.

Finally, the principal-agent problem is a broader category. It arises when one person—the agent—acts on behalf of another—the principal—and their interests or information do not fully align. Moral hazard is one mechanism that can arise within that relationship, but not every delegation produces it.

How to reduce it without giving up protection

No universal mechanism eliminates moral hazard without a cost. Each response distributes risk, responsibility, and the burden of monitoring differently.

Deductibles and copayments leave the insured with part of the loss. Collateral exposes the borrower to a consequence if a project fails. Monitoring makes some decisions easier to observe. Performance-linked contracts bring the interests of the person who delegates closer to those of the person who acts. And credible rules set in advance reduce the expectation that there will always be an extraordinary transfer of costs.

These tools moderate the problem, but do not eliminate it. An excessively high deductible can drain coverage of its value; excessive collateral can exclude creditworthy people who lack assets; and intensive monitoring can consume resources or intrude on areas better left to individual autonomy. Performance incentives can also reward partial metrics while neglecting less visible objectives.

Key idea: Good incentives require balancing two legitimate aims: sharing risks that one person should not bear alone and preserving a meaningful connection between decision and responsibility.

A question of institutional responsibility

Moral hazard helps us look beyond intentions. A program may seek to provide protection, a contract may be voluntary, and aid may respond to an emergency; still, their rules can produce unintended incentives. Analyzing them does not mean denying the value of insurance, lending, or safety nets.

From a liberal perspective, responsibility works best when the person making a decision retains a meaningful share of both benefits and costs. Ownership and responsibility, clear agreements, and predictable rules all support that relationship. But responsibility does not mean abandonment in the face of every loss: it can also be expressed through institutions that allow risks to be shared without guaranteeing that someone else will always pay.

The useful question, then, is not whether all protection makes its recipient irresponsible. It is whether an arrangement can protect people without erasing the signals that guide their behavior. That is the central challenge of moral hazard: designing institutions that distribute losses without completely separating the freedom to decide from the consequences of deciding.