Principal-agent problem

Whenever one person acts on another's behalf, their interests may diverge, and the one in charge cannot fully watch the one doing the work. That is the principal-agent problem.

The principal-agent problem is the difficulty of ensuring that an agent, hired to act on a principal's behalf, actually acts in the principal's interest rather than their own, when their goals differ and the principal cannot fully observe what the agent does. It is one of the most pervasive problems in economics and organisation.

Divergent interests, hidden actions

The problem has two ingredients: the principal and agent want different things, and the principal cannot perfectly monitor the agent. Shareholders, the principals, want managers, their agents, to maximise the firm's value, but managers may prefer perks, empire-building, or a quiet life. Patients want doctors to act in their interest, clients want lawyers to, voters want politicians to, but in each case the agent has their own goals and an informational advantage. Where interests align or actions are fully observable, there is no problem; the difficulty arises when neither holds.

Where it appears

The principal-agent problem runs through the economy. It is at the heart of corporate governance, where the separation of ownership from control means managers may not serve shareholders. It shapes employment, where employers cannot fully observe effort. It pervades the professions, politics, and any relationship of delegation. Agency costs, the losses from this misalignment plus the cost of curbing it, are a real and often large burden.

Managing it

Because the problem cannot be eliminated wherever delegation and hidden action coexist, it is managed by aligning incentives and improving monitoring. Performance-related pay, share options, and bonuses tie the agent's reward to the principal's goals. Monitoring, oversight, audits, and reporting reduce the information gap. Contracts, boards, and reputational stakes all aim to make it in the agent's interest to serve the principal. None is perfect, and each has its own costs and distortions, so the problem is contained rather than solved.

The principal-agent problem is a unifying idea that illuminates corporate governance, employment, politics, and the professions, anywhere one party acts for another. It explains why so much institutional machinery, incentive pay, monitoring, governance, exists to bend the interests of agents toward those they serve, and why the gap between the two can never be fully closed.