11.3 Contracts, principals, and agents
In the last section, we saw that so long as employers do not want to share ownership or profits with their workers, the workers’ incentives will not be aligned with their employers’ incentives.
So how do employers incentivize workers to act in the employers’ interest? To answer that question, we will first look at a critical feature of all employment contracts: their incompleteness.
What is a complete contract?
- complete contract
- A contract is complete if it (a) covers all aspects of the exchange in which any party to the exchange has an interest, and (b) is enforceable (by the courts) at close to zero cost to the parties.
To understand what an incomplete contract is, it is helpful to consider an example of a complete contract. Imagine going to a hardware store and buying a hammer. As the buyer, you expect to pay the price advertised and gain ownership of the hammer you picked out. The seller, the owner of the store, expects to receive an amount of money equal to the price, and in exchange will transfer ownership of the hammer to you.
Although this contract is implicit, it is complete. No aspect of this exchange—the price, the item being exchanged, the transfer of ownership—is excluded from the contract. If one side breaks this contract, say by running out of the store without paying, it is simple and relatively costless to enforce.
Why employment contracts are incomplete
If an employment contract were a complete contract, the employer would be able to specify the number of hours the worker will work, how much effort the worker will exert for each hour, what the worker will do in good or bad work conditions, and so on. Every potential aspect of the employment contract will be specified and laid out in a way that the worker and the employer can understand.
Employment contracts are not the only type of incomplete contracts, but they are especially important in capitalist economies. Other examples of incomplete contracts include insurance, credit contracts (covered in Chapter 17), and lease agreements. Most of the services we buy, such as haircuts, repair work, or healthcare, are also characterized by incomplete contracts.
If the employment contract is complete, then employers could refuse to pay a worker if the worker does not do exactly what the employer wants them to, and courts will enforce the contract quickly and free of charge (at close to zero cost). A complete contract specifies every situation in which the worker’s contract will be terminated if they don’t do what the employer wants them to do. Similarly, if the employer doesn’t pay the worker, then the worker will be able to take the employer to court to ensure they get paid for work they did (quickly and at close to zero cost).
The employment contract, however, can never be complete. There will always be things that the employer and the worker care about that cannot be put into a legally enforceable contract. Here are some key reasons why:
- hidden action problem or moral hazards
- If there is a conflict of interest between a principal and an agent over the agent taking some action that cannot be observed or cannot be verified by a court, then the principal faces a problem of hidden actions. Because the agent’s behavior is hidden, the agent might act in their own interest instead of doing what’s best for the principal, a situation also known as moral hazard.
- non-verifiable information
- Information is non-verifiable if it cannot be verified by a court and therefore cannot be used to enforce a contract.
- asymmetric information
- Information that is relevant to the parties in an economic interaction, but is known by some but not by others.
- The future is unknowable. Not every situation at a job is foreseeable. You cannot write something into a contract that you cannot see coming.
- Effort cannot be directly observed. Even if a firm can afford to monitor a worker’s actions constantly, monitoring effort would require observing the worker’s thoughts and intentions, which is not possible. Worker effort is thus hidden from employers, creating a hidden action problem because the worker can act in their own interest without the employer knowing.
- Effort cannot be the basis of an enforceable contract. Even if an employer could monitor effort, how could they measure it? Information about a worker’s effort is either non-verifiable (an objective third party cannot verify it) or asymmetric (known by some parties to the contract but not others).
- It is difficult to measure work tasks. Many work tasks cannot be clearly or precisely measured (when employees work as part of a team, for example), and therefore they cannot be the basis of an enforceable contract.
- Employment is voluntary. An employment contract cannot prohibit a worker from quitting, as that would not be an employment contract but rather slavery or indentured servitude.
One situation most employers and workers did not foresee was the COVID-19 pandemic. No employment contract could have covered this contingency.
Everyday Economics 11.2
Because effort cannot be measured, companies will come up with substitute measures to make their contracts “more complete” or “less incomplete.” For example, drivers for Uber may have their accounts deactivated “for ratings [on a five-star system] that are below a minimum average rating in their city.” Uber uses passenger and customer ratings as a substitute for “effort” in an attempt to exert greater control over its workers. Can you think of other similar examples from jobs you have held?
The incompleteness of the employment contract can also be a benefit, however, because it allows for flexibility and adaptability.
Principal–agent problems
- principle-agent problem or principle-agent relationship
- A principal-agent relationship or principal-agent problem exists when one party (the principal) would like another party (the agent) to act in some way, or have some attribute, that is in the interest of the principal, and that action or attribute cannot be enforced or guaranteed in a binding contract.
The issue of worker effort is an example of a principal–agent problem. In a principal–agent problem, the principal wants the agent to act in a certain way and/or to have certain attributes. But the principal either does not know or cannot verify whether the agent acts in the desired way or has the attributes they care about. In other words, principal–agent problems are characterized by asymmetric and/or unverifiable information. For the interaction to be a principal–agent problem, the agent’s goals must be different from the principal’s.
A principal–agent problem therefore arises when two conditions hold:
- Conflict of interest: The actions or attributes of the agent affect the principal in such a way that there is a conflict of interest between them.
- Incomplete contract: The agent’s actions or attributes are hidden from the principal (or, if known, they are not verifiable) and so cannot be included in an enforceable contract.
- hidden action problem or moral hazards
- If there is a conflict of interest between a principal and an agent over the agent taking some action that cannot be observed or cannot be verified by a court, then the principal faces a problem of hidden actions. Because the agent’s behavior is hidden, the agent might act in their own interest instead of doing what’s best for the principal, a situation also known as moral hazard.
These conditions can lead to moral hazard, which refers to any situation in which one party to an interaction decides on an action that affects the well-being of the other, but which the affected party cannot control by means of a contract. The term originated in the insurance industry to express the problem that insurers face—namely, the person with homeowner’s insurance may take less care to avoid fires or other damage to their home, thereby increasing the risk above what it would be in the absence of insurance.
The principal–agent problem as a game
We can think of the principal–agent relationship between the worker and the employer as a kind of game. To do so, we need to understand the players, their actions, and their payoffs when they do the best they can.
The players
- The employer is the principal.
- The worker is the agent.
How does each player do the best they can?
- principle of mutual gains and conflicts from exchange
- People mutually benefit by interacting with others but there are typically conflicts over the distribution of these gains.
- conflict of interest
- People have a conflict of interest when they disagree about the outcome of a decision or an allocation. Conflicts of interest over gains from cooperation exist because people disagree about who should get a larger share of the gains.
For the employer, doing the best they can means incentivizing the worker to exert the minimum required effort at the lowest possible cost. For the worker, doing the best they can means maximizing their wage and minimizing their effort. The worker and the employer get mutual gains by cooperating with each other, but they have different goals.
The worker and the employer have a clear conflict of interest because they want opposing things. Add an incomplete employment contract, and a principal–agent problem results.
What actions can each person take?
- The employer chooses the terms of the employment contract: the number of hours to be worked and the wage paid per hour. The employer controls the benefits to the worker of having a job.
- The worker chooses how much effort they will exert for each hour they are employed. Their actions (effort) are hidden from the employer.
- principle of trade-offs and opportunity cost
- The gains you make by choosing some action typically come at the cost of gains that would have been possible had you acted differently.
In choosing their actions, both players face trade-offs. For the employer, paying more will make the worker work harder, but the higher wages increase the employer’s costs. At the same time, workers have to balance the cost of effort (with its corresponding benefit of receiving a wage) against the chance of getting caught if they do not work hard enough.
Who has power in the game?
- bargaining power
- The extent of a person or firm’s advantage in securing a larger share of the economic rents made possible by an interaction.
As we noted in Chapter 10, the game between workers and employers is not a game between equals. Employers have more bargaining power because they make the first move in this game and set the terms of the exchange, which workers then strategically respond to. In this way, this game is similar to the ultimatum game we learned about in Chapter 5.
How do we solve the principal–agent game?
- piece rate
- Under a piece rate contract, a worker is paid a fixed amount (the “rate”) for each unit (“piece”) of the product made.
One option for employers is to pay their workers piece rate, in which pay depends entirely on what a worker produces. In Extension 11.3, we discuss why piece rate is rarely a feasible option.
Recognizing the problem of hidden actions and contractual incompleteness, employers primarily manage the principal–agent problem by making the cost of losing a job high enough that workers work hard to keep that job.
Extension 11.3 Why don’t employers pay piece rate?
Workers shucking clams in 1983 in Bivalve, New Jersey. The workers were paid piece rate, based on how many clams they shucked.
Everyday Economics 11.3
Think back to the examples of the prisoners’ dilemma, the public goods game, and the common pool resource problem in Chapters 6 and 7. How is the hard-to-measure output of a team like these examples? What characteristics of a team’s output make it hard to understand what an individual member contributes? What experiences have you had working on teams at school, college, or a job?
In the late 1800s, over half of manufacturing workers in the United States were paid by piece rate. Today, however, piece rate pay is rare, for a few reasons:
- It is difficult to measure output. Think about someone employed as an administrative assistant, nurse, or accountant. How would you go about measuring their output? What would you measure? How would you value that output financially?
- Output is not always tied to effort. Consider a checkout clerk at a store. Although their output can be easily measured, how many items they ring up is largely out of their control.
- Most people work in teams. Rewarding members of a team for the team’s output can create a free-rider problem or create tension within the team if there is disagreement about how much each person contributed.
Even for those jobs where piece rate pay is appropriate, it is only a partial solution. One problem for employers is that workers getting paid piece rate are incentivized to focus on quantity over quality of production. Low quality could hurt the firm’s profitability if it makes the firm’s products less popular, which could be bad for both the employers and the workers. But it is hard, and maybe even impossible, to measure quality exactly. There’s no reliable way to ensure that workers are paid only for work of a certain quality.
- technological progress
- A change in technology that reduces the amount of resources (labor, machines, land, energy, time) required to produce a given amount of the output.
- technological progress
- A change in technology that reduces the amount of resources (labor, machines, land, energy, time) required to produce a given amount of the output.
- learning by doing
- People learn better (less costly) ways of working by developing individual skills and discovering better ways to organize production among members of a team.
That piece rate is only a partial solution was further demonstrated in an experiment by the economists David Atkin, Azam Chaudry, Shamyla Chaudry, Amit Khandelwal, and Eric Verhoogen. They conducted this experiment in 2014 in Sialkot, Pakistan, where about 40% of the world’s soccer balls are produced, including soccer balls used in some World Cup and Olympic events. The economists had invented a new technology for cutting the material that goes into soccer balls. This new technology reduced the total cost of production and wasted less material—it was a clear case of technological progress. However, despite offering this technology to many firms for free, few of the manufacturers adopted the technology.
The vast majority of the workers making soccer balls were paid piece rate. Although the workers knew the advantages of the new technology, they also knew that the process of learning-by-doing would have, at least initially, slowed down their daily production, meaning they would make less money. Because the workers would lose money in the short run and not expect to receive any of the added profits this new technology might bring, they had no reason to adopt it. Many of these workers thus lied to their employers about the value of the technology. In sum, the incentives of the workers were not fully aligned with the incentives of their employers.
Exercise E11.4 Why piece rates don’t work for every job
Using the same job titles as in Exercise 11.5, explain why it would be difficult or impossible to compensate a worker with that job title using piece rate.
Question E11.4
Which of the following statements about piece rate pay are correct? Choose all that apply.
- In practice, piece rate workers may resist new technologies if adoption slows short-run output and reduces their pay.
- Team production makes it hard to link pay fairly to individuals.
- Piece rate is used when effort is not directly observable, but output is.
- Piece rate does not fully align incentives; workers may resist new technology or focus only on quantity.
Exercise 11.4 Conditions for a principle–agent problem
Two conditions are necessary for a principal–agent problem to exist: conflict of interest and an incomplete contract. Why are both of these conditions necessary for something to be a principal–agent problem?
Exercise 11.5 Incomplete employment contracts
The following table lists three job titles on the left. For each job title, fill in the blank spaces in that row, which describe why employment contracts for that job are incomplete. For the bottom row, use a job title that you have had or are interested in having, then fill in the rest of the row as you did for the other three.
| Job title | Example of a task or responsibility | Why the contract is incomplete for this task | Why this task is hard to enforce or measure |
|---|---|---|---|
| Nurse | |||
| Software developer | |||
| Firefighter | |||
Question 11.3
Which of the following are reasons why employment contracts are incomplete? Choose all that apply.
- Employment contracts cannot prevent workers from quitting, because that would be indentured servitude or slavery, which are not legally enforceable.
- The future is unknowable, and not every situation that may arise on the job can be anticipated or written into a contract.
- Worker effort is hidden, often non-verifiable or asymmetric, and so it cannot be the basis of an enforceable contract.
- “Unfinished” is not the reason contracts are incomplete. Instead, the problem lies in enforceability, hidden information, and unforeseeable situations.
Question 11.4
Which of the following relationships may result in a principal–agent problem? Choose all that apply.
- The lender (principal) provides funds but cannot perfectly monitor how the borrower (agent) will use them, creating risk of misuse.
- The gas station–customer relationship is a straightforward market exchange. Both sides get what they pay for at the point of transaction, with little hidden action.
- The firm (principal) hires the worker (agent) but cannot directly observe or contract for effort, creating a principal–agent problem.
- The client (principal) hires the lawyer (agent) but cannot perfectly observe the quality or amount of effort provided, creating a principal–agent problem.
-
Martin Brown, Armin Falk, and Ernst Fehr. 2004. “Relational Contracts and the Nature of Market Interactions”. Econometrica 72(3): pp. 747–780. ↩
