Loading field map
Contract theory is the branch of microeconomics that studies how parties design agreements when information is imperfect and actions are difficult to verify. Its central subject is the contract: a formal or informal arrangement in which one party (the principal) delegates a task or transfers resources to another (the agent), specifying what each side will do, receive, or bear under various contingencies. The field asks why contracts take the shapes they do, what problems they solve, and what inefficiencies remain when they cannot solve them.
The stakes are broad because contracts are the elementary particles of economic exchange. Employment relationships, insurance policies, credit agreements, franchise arrangements, and government procurement all involve one party acting on behalf of another under conditions of hidden information or hidden action. Contract theory provides the analytical vocabulary for understanding these arrangements: incentive constraints, information rents, moral hazard, adverse selection, and incomplete contracting.
The foundational premise of contract theory is that the parties to an agreement have conflicting interests and that the agent possesses information or control that the principal does not. If both parties observed everything and could verify every action in a court, contracting would be trivial: the parties could simply write a complete plan of action and enforce it. The field's entire apparatus exists because this condition fails in three distinct ways.
Hidden information (adverse selection) arises when the agent knows something relevant before the contract is signed—their own ability, the quality of the good they sell, the riskiness of their project—that the principal cannot observe. The principal's problem is to design a menu of contracts that induces the agent to reveal this information truthfully, or at least to act in a way that respects it.
Hidden action (moral hazard) arises after the contract is signed, when the agent chooses an effort level or a risk profile that the principal cannot observe. The principal cannot directly command the desired action; they must instead create incentives through the payment structure, knowing that the agent will respond to those incentives in their own interest.
Hidden information about outcomes (verifiability problems) arises when the agent observes the outcome of their actions but the principal cannot confirm it to a third party. Even if both parties know what happened, a court may not be able to verify it, making contracts contingent on that outcome unenforceable.
These three problems are not mutually exclusive; real contracts typically involve all of them. A manager hiring a salesperson faces adverse selection (does the candidate have talent?), moral hazard (will they work hard?), and verifiability (can the firm prove how much effort contributed to sales?). Contract theory's power lies in isolating each problem and showing how the optimal contract responds to it.
The workhorse of the field is the principal–agent model, a formal framework in which a principal offers a contract to an agent, the agent accepts or rejects, and then the agent takes an action that affects both parties' payoffs. The principal's problem is to choose the contract that maximizes their own expected payoff subject to two constraints.
The participation constraint (or individual rationality constraint) requires that the agent be at least as well off accepting the contract as taking their best outside option. If the contract does not meet this bar, the agent walks away.
The incentive compatibility constraint requires that the agent, choosing among all available actions, prefers the action the principal wants to induce. The contract must make the desired action the agent's best response.
The central result of this framework is that when the agent is risk-averse and effort is unobservable, the optimal contract involves a trade-off between insurance and incentives. A fixed wage would fully insure the agent against income risk but provide no incentive to work. A contract that pays entirely based on output would provide strong incentives but force the risk-averse agent to bear income risk they dislike. The optimal contract balances these forces: the agent's pay is partially tied to output, but not fully, and the agent earns an information rent—extra compensation beyond their reservation utility—because the principal cannot extract all the surplus when the agent's information or actions are private.
This framework yields a set of robust comparative statics. Incentive intensity should be higher when the agent is less risk-averse, when effort has a larger effect on output, and when output is a less noisy signal of effort. The framework also explains why contracts often bundle multiple performance measures, why tasks that are difficult to measure get under-incentivized, and why agents are often given a choice among contracts that screen their types.
When the agent has private information before contracting, the principal's problem shifts from motivating effort to extracting information. The classic setting is a monopolist selling to customers with different willingness to pay, or an insurer offering policies to customers with different risk types. The principal cannot observe the agent's type, but can offer a menu of contracts designed so that each type chooses the contract intended for them.
The key concept is the revelation principle: any outcome achievable through a complex negotiation can be achieved by a direct mechanism in which agents report their type and receive an allocation designed for that report, provided the mechanism is incentive compatible. This principle simplifies analysis enormously, allowing the theorist to focus on truthful reporting without loss of generality.
The optimal screening contract typically involves distortions away from the first-best. The principal must give the most efficient or lowest-risk type an information rent to prevent them from mimicking less attractive types. To reduce this rent, the principal distorts the contracts offered to lower types—for example, by reducing the quantity sold to low-valuation customers or by offering incomplete insurance to low-risk customers. The result is a menu of contracts that sorts types by their choices, with the highest type receiving an efficient allocation and lower types receiving progressively more distorted ones.
A related concept is signaling, which reverses the direction of information flow. Instead of the uninformed party designing a menu, the informed party takes a costly action to reveal their type. Education as a signal of ability, warranties as signals of product quality, and dividend payments as signals of firm profitability are canonical examples. The Spence signaling model shows that for a signal to be credible, it must be more costly for low types to produce than for high types—otherwise everyone would send it and it would convey nothing.
When the agent's action is unobservable, the principal faces the problem of designing incentives for effort. The basic model assumes the agent chooses effort, output is a stochastic function of effort, and the agent is paid according to the realized output. The optimal contract ties pay to output, but the strength of the tie depends on the parameters discussed above.
Several extensions deepen the analysis. Multitasking shows that when the agent performs several tasks and only some are measurable, paying strongly for the measurable tasks diverts effort away from the unmeasurable ones. This explains why teachers are not paid solely on test scores, why salespeople are not paid solely on volume, and why firms use subjective evaluation alongside objective metrics.
Career concerns models show that even without explicit incentive contracts, agents work hard because their current performance affects their future wages. The agent's concern for their reputation substitutes for formal incentives, and the optimal contract may be flatter than it would be without this implicit motivation.
Relational contracts extend the analysis to settings where formal enforcement is impossible. If the parties interact repeatedly and both value the future, they can sustain cooperative arrangements in which each side performs because the threat of ending the relationship disciplines both. These informal agreements are widespread in firms, where many obligations are too subtle or too costly to specify in enforceable terms.
A major branch of contract theory departs from the assumption that contracts can specify everything relevant. Incomplete contract theory starts from the observation that real contracts are grossly incomplete: they leave many contingencies unspecified, they are vague about quality, and they are costly to renegotiate. The question is not how to design the optimal complete contract, but what parties do when they cannot write one.
The most influential answer comes from the property rights approach to the firm, developed by Oliver Hart and his collaborators. If contracts cannot specify what happens in every future state, then the allocation of residual control rights—the right to make decisions in circumstances not covered by the contract—matters. The owner of an asset has the right to decide its use when the contract is silent. This ownership allocation affects the incentives of the parties to invest in relationship-specific assets, because an investor who does not own the asset fears that the other party will expropriate the value of their investment during renegotiation.
This framework explains the boundaries of the firm. When two parties must make complementary investments and contracts are incomplete, merging them under common ownership may improve incentives by giving each party more control over the assets they need. But integration also has costs: the party who loses ownership of an asset loses the incentive to invest in it. The optimal ownership structure balances these forces. This approach has been used to analyze vertical integration, joint ventures, privatization, and the allocation of control within organizations.
The incomplete contracts approach is methodologically distinct from the complete-contract tradition. It does not derive optimal contracts from first principles; instead, it takes incompleteness as a primitive and asks how ownership and governance structures compensate for it. Critics have argued that the approach lacks a rigorous foundation for why contracts are incomplete, and that its predictions depend on assumptions about renegotiation that are not always spelled out. Defenders respond that the approach captures an important reality that complete-contract models cannot: the role of authority and control in economic life.
Contract theory emerged as a distinct subfield in the 1970s, though its intellectual roots reach back further. The principal–agent problem was anticipated in early work on risk-sharing, which asked how risk should be allocated between parties with different attitudes toward risk. The modern field crystallized when economists began asking not just how risk should be shared, but how the information structure of the relationship shapes what can be achieved.
The 1970s and 1980s saw the formalization of the core models. The adverse selection framework was developed through work on regulation and on insurance markets, where the problem of sorting different risk types was particularly salient. The moral hazard framework was formalized through the analysis of sharecropping, insurance with deductibles, and managerial compensation. The revelation principle, which unified much of this work, was established in the late 1970s and became a standard tool.
The 1980s and 1990s brought the incomplete contracts revolution, which shifted attention from the design of optimal payment schemes to the allocation of control rights. This work connected contract theory to the theory of the firm and to corporate governance, and it remains influential in those areas.
A parallel development was the rise of mechanism design, a closely related field that asks how institutions can be designed to achieve desired outcomes when information is dispersed. Mechanism design is sometimes treated as a sibling of contract theory, sometimes as a generalization of it. The distinction is one of emphasis: mechanism design focuses on the design of rules for collective decision-making, while contract theory focuses on bilateral relationships and the allocation of risk and incentives. In practice, the two fields share tools and results, and many contributions belong to both.
Contemporary contract theory is a mature field with a well-established toolkit. The core models are taught in graduate microeconomics courses and are used as building blocks in applied fields: labor economics uses moral hazard models to study compensation; corporate finance uses adverse selection models to study capital structure; industrial organization uses screening models to study price discrimination; public economics uses contract theory to study regulation and procurement.
Several developments characterize the current landscape. Behavioral contract theory incorporates psychological findings—loss aversion, fairness concerns, present bias—into contract models, asking how optimal contracts change when agents are not fully rational. This work has produced new predictions about the prevalence of flat wages, the role of reference points in employment relationships, and the design of commitment devices.
Dynamic contract theory extends the static models to repeated relationships, asking how incentives and information evolve over time. This work has illuminated the design of long-term employment contracts, the role of tenure and promotion, and the dynamics of reputation.
Contract theory with limited commitment relaxes the assumption that parties can commit to long-term agreements, asking what happens when either side can walk away or renegotiate. This has been important for understanding sovereign debt, labor markets with turnover, and the limits of insurance.
The field also faces ongoing methodological debates. The complete-contract tradition is sometimes criticized for assuming unrealistic levels of sophistication on the part of the contracting parties, while the incomplete-contract tradition is criticized for lacking a fully rigorous account of why contracts are incomplete. These debates are productive: they have generated a large literature on the foundations of contract theory, including work on the costs of writing contracts, the limits of verifiability, and the role of courts and legal institutions.
A notable feature of the field is its close relationship with law and economics. Contract theory provides a framework for understanding why legal rules matter, how courts should interpret ambiguous contracts, and what the limits of private ordering are. The field has also influenced the design of real institutions, from the structure of executive compensation to the design of procurement auctions and the regulation of utilities.
The field's central insight remains its most durable contribution: that the terms of an agreement are not just a division of surplus but a response to the information and incentive problems that plague all exchange. Contracts are solutions to problems, and the shape of the solution reveals the structure of the problem. This perspective has proven remarkably portable, and it continues to organize how economists think about organizations, markets, and the design of institutions.