Principal–Agent Problems: Why Can Someone Acting for You Choose Differently?

A principal–agent problem does not mean that an agent is necessarily dishonest. It arises when one person asks another to act on their behalf, their aims are not perfectly aligned, and the principal cannot observe every action. The agent usually has more information and faces different costs, so choices may diverge from the principal’s interests even when the agreement is clear.

Consider an owner who appoints an agent to lease an apartment. The owner cares about a stable tenant, proper maintenance and long-term returns. The property manager may care about these too, yet may place more weight on signing quickly because repeated inspections consume time and the fee is usually earned after a lease is secured. The issue is not that either party is “bad”; the same decision carries different costs and benefits for each, while the manager knows more about the inspection process.

This differs from delegation itself. Delegation describes how authority moves; the principal–agent problem asks whether information and incentives still direct that authority towards the original goal. Responses may include verifiable outcomes, better access to information, adjusted remuneration and review mechanisms. Yet monitoring costs money, and narrow metrics can encourage attention only to what is measured. The central point is that handing over a task does more than distribute authority: it redesigns the relationship among information, responsibility and motivation.

Source: https://www.imf.org/en/publications/wp/issues/2016/12/31/supervisory-incentives-in-a-banking-union-44273