A homeowner with comprehensive flood insurance has less incentive to invest in flood-proofing her basement than a homeowner who pays out of pocket for flood damage. Fully covered, the incremental value of prevention falls to zero — the insurer bears the cost of damage either way. The homeowner hasn't become dishonest; she has simply responded rationally to changed incentives. Her risk-taking behavior has shifted because the cost of the risk is now borne by someone else. That behavioral change — driven by the transfer of risk — is moral hazard.
In plain terms
Moral hazard is the tendency of a party insulated from risk to behave differently — typically taking more risk or exercising less care — than they would if they bore the full consequences of their actions. It arises after a contract or transaction that shifts risk from one party to another, in contexts where the risk-bearing party cannot fully observe or control the behavior of the risk-taking party.
The "moral" label is historical — it implies irresponsible behavior — but economists use it neutrally to describe a rational response to changed incentives. Anyone facing a reduced private cost of risky behavior will rationally accept more risk. This is not moral failure; it is predictable optimization under the new incentive structure.
Moral hazard requires two conditions:
- Risk transfer: one party bears the cost of outcomes they don't fully control (insurer pays claims; lender loses principal on bad loans; government guarantees bank deposits).
- Hidden action: the party whose behavior creates risk can't be fully observed or monitored by the party bearing the risk.
Why it works this way
Insurance theory illustrates the mechanism clearly. Without insurance, a firm's risk-prevention investment is determined by the expected cost of losses multiplied by their probability — the full actuarial case for caution. With full insurance at a flat premium, the expected private cost of losses is zero — the insurer absorbs all claims. The marginal value of loss-prevention falls to zero. Risk-prevention investment falls accordingly.
The National Flood Insurance Program's data on repetitive loss properties documents moral hazard in federal flood insurance: properties that have flooded multiple times — and been fully paid out multiple times — are often rebuilt identically rather than elevated or flood-proofed. Subsidized insurance with below-actuarial premiums removes the incentive to adapt, and the government bears repeated losses on properties whose owners face little marginal cost of flood damage.
A real example
The 2008 financial crisis involved severe moral hazard in banking. Large banks that operated with implicit government guarantees ("too big to fail") took on more leverage and risk than they would have if they expected to bear the full cost of failure. The Federal Reserve's post-crisis regulatory framework — capital requirements, living wills, and stress tests — is explicitly designed to reduce moral hazard by making banks bear more of the cost of their own failure.
Why it matters
Moral hazard is why insurance contracts have deductibles, coinsurance requirements, and policy limits. It is why employment contracts link pay to performance. It is why bank regulation requires capital cushions. Every mechanism for managing principal-agent and insurance relationships exists because pure risk transfer produces behavioral responses that increase the risk being insured. Aligning incentives — maintaining some skin in the game for the risk-taker — is the universal tool for managing moral hazard.
