Skip to content
The Chat Archive
o3-2025-04-16· OpenAI LMArena Human Preference 140K

What's moral hazard in contract theory?

  1. User

    What's moral hazard in contract theory?

  2. Assistant

    Moral Hazard — A Quick Primer in Contract Theory

    Moral hazard is a problem that arises after two parties sign a contract when:

    1. One party’s actions are unobservable (or imperfectly observable) to the other.
    2. Those hidden actions affect the outcome that both parties care about.
    3. Because the actions can’t be perfectly monitored, the acting party may behave differently from what was agreed, shifting some risk or cost onto the other party.

    Put differently, the agent (the one taking the hidden action) has an incentive to shirk, take excessive risk, or otherwise deviate from the effort level promised, because the principal (the one paying for the result) can’t perfectly detect or punish that deviation.


    Classic Examples

    SettingPrincipalAgentHidden Action / RiskResulting Hazard
    Auto insuranceInsurance companyDriverHow carefully the driver drivesDriver might drive less carefully (e.g., speed more) once insured.
    EmploymentFirm/ShareholdersEmployee/ManagerEffort exerted on the jobWorker may slack off if effort isn’t monitored.
    FinanceTaxpayers“Too-big-to-fail” bankLevel of portfolio riskBank might take on more risk, expecting a bailout if things go south.

    Why It Matters

    • Inefficient effort level: The agent may supply less effort than is socially optimal.
    • Risk shifting: The agent may take on riskier projects than the principal would want.
    • Welfare loss: Resources get misallocated, causing deadweight loss in the overall economy.


    Contract-Theory Toolkit to Mitigate Moral Hazard

    1. Performance-based pay
      Align incentives by tying compensation (bonuses, commissions, stock options) to observable outcomes.

    2. Monitoring & reporting
      Introduce audits, reports, GPS trackers, or periodic reviews to make hidden actions more observable.

    3. Skin in the game
      Require the agent to bear some downside risk (deductibles, equity stakes, collateral).

    4. Efficiency wages & career concerns
      Pay above-market wages that the agent stands to lose if caught shirking, or rely on reputation effects.

    5. Termination clauses & penalties
      Threaten firing or financial penalties if performance drops below target levels.

    6. Team incentives & peer monitoring
      Use group bonuses so peers police one another (common in law firms, consultancies).


    A Sketch of the Formal Model (very high-level)

    1. Principal wants to maximize expected utility:
      max⁡w(e)  E[V(e)−w(e)]\max_{w(e)} \; \mathbb{E}[V(e) - w(e)]
      where (e) = effort, (V(e)) = value generated, (w(e)) = wage/payment.

    2. Agent chooses effort to maximize their expected utility:
      max⁡e  E[w(e)]−C(e)\max_{e} \; \mathbb{E}[w(e)] - C(e)
      subject to (i) Incentive-compatibility (IC) and (ii) Participation (IR) constraints.

    3. Because (e) is unobserved, the contract must make the agent’s best-response effort coincide with the principal’s desired effort, given risk preferences and outcome noise. This usually implies:
      – The principal must leave some information rent (extra surplus) to the agent.
      – First-best (full information) effort is generally not attainable.


    Bottom Line

    Moral hazard in contract theory is all about hidden actions after a contract is in force. The challenge is designing contracts or institutional arrangements so that the agent’s private incentives line up with the principal’s goals, minimizing welfare losses without imposing excessive risk or monitoring costs.

Responses were produced in different, uncontrolled contexts (system prompts, settings, prior turns). This is not a benchmark.