<- Back to laws

Contracting problem under asymmetric information

Principal-Agent Problem

When one party delegates work to another but cannot fully observe action, information, or objectives, a contract must trade stronger incentives against risk, measurement error, and distorted behavior.

Scientific statusEconomic modeling framework
Predictive formOptimization under information asymmetry
DomainDelegation and contracts
EvidenceTheory + field and laboratory studies
Key limitationModel and measurement assumptions
Common misuseMore performance pay always aligns interests
INTERACTIVE MODEL

contract = fixed pay + beta x measured performance

In a simple linear contract, beta is incentive intensity. A larger beta can make effort more valuable to the agent, but also transfers performance risk and rewards whatever the measure captures rather than the principal's full objective.

The contract table lets measurement quality vary while holding ability, effort cost, outside options, and risk preferences fixed. Its effort, risk, and distortion scores are teaching indices, not an optimal-contract calculation.

48.0Illustrative effort index
(index)
0 %100 %
CONTRACT TENSION TABLEWhat gets rewarded pulls effort through the measurement aperture.
PRINCIPALFULL OBJECTIVE
QUALITYSPEEDTRUSTRESILIENCE
CONTRACTBETA 40%
AGENTALLOCATED EFFORT
MEASURED TASK
UNMEASURED WORK
MEASUREMENT APERTURE
EFFORT0RISK BORNE0DISTORTION0UNMEASURED CAPACITY0

Increase performance pay to strengthen the measured beam. Then narrow the aperture: effort can rise while the broader objective receives less attention.

CHANGE
Performance-pay share
WATCH
effort + risk
MEANING
The contract table lets measurement quality vary while holding ability, effort cost, outside options, and risk preferences fixed. Its effort, risk, and distortion scores are teaching indices, not an optimal-contract calculation.
VISUAL MODEL

Delegation creates a gap between the goal and the signal.

The principal values a broad outcome, while the contract can reward only an observed measure. Stronger incentives pull effort toward the measured target but can leave unmeasured work behind.

principal objectivemeasured signalagent action
01 / MEANING

What it actually says

A principal-agent relationship exists when a principal delegates an action or decision to an agent. The problem becomes economically interesting when their objectives differ and the principal cannot costlessly observe the agent's action, private information, or the state of the world. The agent may then choose behavior that is individually rational but not best for the principal.

There is no single principal-agent law and no universal compensation formula. Agency theory is a framework for designing contracts, monitoring, ownership, authority, information, and governance. Moral hazard concerns hidden action after contracting; adverse selection concerns hidden information or type before or during contracting. Professional analysis keeps these distinct.

Compact formcontract = fixed pay + beta x measured performance
Best interpretationDelegation and contracts evidence in incentives.
Important cautionModel and measurement assumptions.
"A useful law compresses a pattern. It does not erase the conditions that make the pattern true."
02 / ORIGIN

How the idea developed

The modern form emerged through observation, argument, and later refinement. The timeline separates the first insight from the version now used in textbooks and practice.[1]

19731973

Stephen Ross formalizes the principal's problem as an economic theory of agency.

19761976

Jensen and Meckling define agency costs and connect them to ownership, monitoring, and the firm.

19791979

Holmstrom characterizes when additional performance information improves a moral-hazard contract.

TodayToday

Contract theory studies multitasking, teams, careers, relational contracts, mechanism design, regulation, platforms, and algorithmic management.

Historical cautionEponymous laws often change after their first publication. Popular wording may be broader and cleaner than the original evidence.
03 / MECHANISM

How the pattern works

The relation becomes useful only when its mechanism, measurement process, and operating range are visible.

01Goal divergence

The principal and agent value outcomes, effort, risk, time, or private benefits differently.

02Information asymmetry

Action, type, conditions, or output quality cannot be observed or verified perfectly.

03Incentive-risk trade-off

High-powered pay can encourage effort while imposing noise the agent may demand compensation to bear.

04Measurement substitution

The agent rationally shifts effort toward rewarded measures, potentially neglecting quality, maintenance, cooperation, or long-term value.

MODELcontract = fixed pay + beta x measured performance

In a simple linear contract, beta is incentive intensity. A larger beta can make effort more valuable to the agent, but also transfers performance risk and rewards whatever the measure captures rather than the principal's full objective.

04 / APPLICATIONS

Where it earns its keep

Applications are strongest when the law changes a decision, measurement, model, or experiment rather than merely providing an analogy.

CORPORATE GOVERNANCE

Design oversight and executive contracts

Application

Boards combine compensation, monitoring, ownership, disclosure, and decision rights to constrain agency costs.

PROFESSIONAL NOTE

Pay metrics should match controllable long-term value and account for risk, manipulation, and horizon.

PUBLIC SERVICES

Avoid single-metric performance regimes

Application

Schools, hospitals, policing, and regulation contain multidimensional goals that one score cannot fully represent.

PROFESSIONAL NOTE

Use balanced evidence, professional norms, audits, and outcome review instead of one high-powered target.

DIGITAL PLATFORMS

Govern delegated algorithms and vendors

Application

Advertisers, users, platforms, creators, and model providers delegate decisions across several layers.

PROFESSIONAL NOTE

Map each principal, agent, information gap, and externalized cost; the relationship is rarely one simple pair.

05 / LIMITS & MISUSE

Where it stops working

Canonical models often assume known preferences, rational optimization, contractible outputs, stable technology, and a clear principal. Real organizations contain multiple principals, teams, intrinsic motivation, power, identity, incomplete contracts, legal constraints, and contested objectives.

Observed low effort does not prove opportunism. Bad tools, ambiguity, overload, missing capability, unfairness, conflicting principals, or an impossible target can produce the same signal. Monitoring itself is costly and can crowd out trust or redirect attention.

Misuse

"Employees are naturally untrustworthy"

Better: Agency problems arise from structure and information, not a universal moral defect.
Misuse

"More variable pay solves alignment"

Better: It can increase risk, gaming, short-termism, and neglect of unmeasured work.
Misuse

"The shareholder is the only principal"

Better: Organizations often answer to several stakeholders, authorities, and legal duties.
Misuse

"Monitoring can eliminate agency cost"

Better: Observation is incomplete, costly, and may change behavior in undesirable ways.
07 / REFERENCES

Sources and further reading

Original publications and serious secondary scholarship are prioritized over summaries.

  1. Ross - The Economic Theory of Agency: The Principal's ProblemThe foundational 1973 formulation from the principal's perspective.https://www.aeaweb.org/aer/top20/63.2.134-139.pdf
  2. Jensen and Meckling - Theory of the FirmDefines agency relationships and agency costs in ownership and finance.https://doi.org/10.1016/0304-405X(76)90026-X
  3. Holmstrom - Moral Hazard and ObservabilityClassic result on using additional information in incentive contracts.https://doi.org/10.2307/3003320
  4. Prendergast - The Provision of Incentives in FirmsBroad empirical and theoretical review of organizational incentive systems.https://doi.org/10.1257/jel.37.1.7
CONTINUE EXPLORING

Related laws, with the relationship made explicit.

These are editorial connections, not claims that the laws are mathematically equivalent.

CONTINUE READING

Place this law inside the collection.

LAW 034 / 100 PUBLISHED