Chapter 23: Asymmetric Information

Policy Example: Should the Government Mandate the Purchase of Health Insurance?

A key component of the Affordable Care Act, passed by congress and signed into law by President Obama in 2010, was the individual mandate. This compelled individuals to purchase health insurance or face a large financial penalty. This provision proved to be the most controversial of the new law, yet many economists believed it to be the crucial piece that guaranteed the success of the new law.

Exploring the Policy Question

1. Why do many economists think it essential that the individual mandate be a part of the Affordable Care Act?
2. Is the individual mandate the only way to achieve universal health insurance? What other ways do societies accomplish this?

23.1 The Market for Lemons Problem

Learning Objective 23.1: Describe the lemons problem in markets with asymmetric information.

23.2 Adverse Selection

Learning Objective 23.2: Explain the term adverse selection and how it affects insurance markets.

23.3 Principal-Agent Models

Learning Objective 23.3: Describe how asymmetric information can affect principal-agent relationships.

23.4 Moral Hazard

Learning Objective 23.4: Explain moral hazard and how it can affect the efficiency of markets.

23.1 The Market for Lemons Problem

Learning Objective 23.1: Describe the lemons problem in markets with asymmetric information.

One of the assumptions of the efficient markets hypothesis is that buyers and sellers are completely informed, they know everything there is to know about the goods for sale in a market, like their quality. This is critical for achieving efficiency since a buyer’s willingness to pay for a good depends on them knowing the value of the good to themselves. If a buyer, for example, over estimates the value of a good to themselves, perhaps by over-estimating its quality, then they might end up buying it at a price that exceeds their willingness to pay for the good once its quality is revealed. This would be an exchange that lowers total surplus and the market cannot therefore be efficient.

Asymmetric Information describes a situation when one side of an exchange, the buyer or the seller, knows more about the product than the other. Generally, we might expect the sellers of goods to know more about them than buyers, but not always – a collector of antiques might know more about the value of an item they see for sale by someone who found an old item in their attic. Items whose value is not immediately known to a buyer or seller are said to have hidden characteristics. Asymmetric information can also arise when agents’ actions are not visible to all parties. For example, a customer may hire a mechanic to fix their car, but they do not observe the actions the mechanic actually takes. The mechanic might say they replaced a part when, in fact, they did not. We call these hidden actions.

It situations where both parties to a transaction have the same information, either full or partial information, neither party has an advantage over the other. However, when one party has access to more information this can lead to opportunistic behavior, where the more informed party takes advantage over the less informed party for economic gain. This advantage taking leads to market failures where transactions fail to yield efficient outcomes. Two main problems associated with asymmetric information are Adverse Selection and Moral Hazard. Adverse selection refers to the situation where asymmetric information on the part of one party in an economic transactions leads to desirable good remaining unsold, even though they would be sold in a market with full information.

The classic example of this is called the market for lemons after George Akerlof’s 1971 paper of the same name. In this scenario, sellers of used cars possess more information about the true quality of the cars than sellers do since the owners of used cars have been using them and know their faults. In a simplified version of this model, consider a world in which there are only two types of cars: good quality cars and bad quality cars (or ‘lemons,’ which in the United States is slang for a bad car). There are equal numbers of both in the marketplace and both buyer and sellers know this. Buyers value good used cars at $10,000 and lemons at $5,000. Sellers are willing to sell good used cars for $8,000 and lemons for $3,000. For clarity, let’s suppose that there are exactly 100 cars of each type and over 200 buyers, each willing to buy either car, given the right price.

Immediately we can see that in a world of full information an efficient outcome will arise: owners of good cars will sell them at a price between $8,000 and $10,000 and owners of lemons will sell them at a price between $3,000 and $5,000. Each car sold will generate a total of $2,000 in total surplus regardless of the price agreed to as this is the difference between the sellers’ minimum willingness-to-accept and the buyers’ maximum willingness-to-pay. In the end the sake of 200 cars will yield a total surplus of $400,000 (or 200 cars times $2,000 surplus for each).

Now consider a world of asymmetric information where the sellers know the quality of the cars they are selling, and the buyers do not. Sellers of both types of cars have an incentive to claim that their cars are of good quality. Buyers understand the incentive to misrepresent the true quality of the car and therefore do not believe the sellers claims. So what happens in the market? Let’s keep things simple by assuming that buyers are risk neutral. In this case the buyers know that, with equal amounts of both types of cars in the market that choosing one at random will yield and expected value of $7,500.

Probability of a good car = .5, value of a good car = $10,000.

Probability of a bad car = .5, value of a lemon = $5,000.

Expected value = (.5)($10,000) + (.5)($5,000) = $7,500

This means that no buyer is willing to pay more than $7,500 for a used car. Since this is true no seller of a good car is willing to sell as $7,500 is lower than their minimum willingness-to-accept. Because of this, no owner of a good quality used car will offer them or sale and the only cars for sale on the market will be the lemons. Both buyers and sellers can figure this out son in the end buyers know that only lemons are for sale and will not offer more than $5,000. This leaves a market for only lemons in which all 100 lemons will be sold for a price between $5,000 and $3,000. Each transaction generates $2,000 in total surplus for a total of $200,000. This is the market failure: the asymmetric information problem leads to a deadweight loss of $200,000 or the difference between the total surplus in the full information marketplace and the total surplus in the market with asymmetric information.

The fact that the good quality cars disappear from the market is called Adverse Selection, a topic we will focus on in the next section.

23.2 Adverse Selection

Learning Objective 23.2: Explain the term adverse selection and how it affects insurance markets.

Adverse Selection occurs when the more desirable attributes of a market withdraw due to asymmetric information. This could be better quality products, better quality consumers or better quality sellers. The result is that consumer and producers may not make transactions that are socially beneficial, transactions that would yield positive producer and/or consumer surplus. This is the market failure associated with asymmetric information: the ability of the more informed agents to exploit their advantage.

Insurance markets are a classic example of market failures from information asymmetries. Information asymmetries exist in insurance markets due to the fact that there is both hidden action and hidden characteristics: insurers cannot monitor the actions and private information of the insured. For example, in car insurance, the insurer does not know how carefully a driver actually drives. In health insurance, the insurer might not know about pre-existing conditions and the lifestyle of their insured. Because of this information asymmetry insurers have to charge a price that is an average of the costs of their insured. This price is often too high for the most desirable consumers causing them to not purchase insurance. This decision to stay out of the market makes the situation worse as the average cost of the insured increases as the lowest cost consumers exit, further raising the price of insurance forcing even more of the healthier consumers from the market.

This outcome is inefficient because if insurance companies had full information about their clients, they could charge each a price that is above the cost of insurance but is less than the risk premium – the amount above the cost that consumers are willing to pay to avoid risk – that would allow these healthier clients to purchase insurance and create more surplus in society.

23.3 Principal-Agent Models

Learning Objective 23.3: Describe how asymmetric information can affect principal-agent relationships.

Principal-agent relationships are situations in which one person, the principal, pays another person to perform a task for them. In its most basic form, this describes the employee-employer relationship. But is can also describe a situation in which a car owner pays a mechanic to fix their car, or a homeowner hires a housecleaner or many other everyday situations. These relationships are often subject to asymmetric information because agents can act in ways that are unobserved by the principal. If the individual incentives are not aligned, the agents might take actions contrary to the interests of the principal. For example, the owner of a car in need of repair would like the car to be repaired properly and inexpensively as possible. The mechanic’s incentives might be to repair the car properly, but to maximize their revenue from the situation. So, for example, the mechanic might like to perform extra, unnecessary work in order to earn more money. Since the car owner does not know that the extra work is unnecessary, the mechanic can claim it is and earn the extra revenue. Similarly, the employers incentive is to have each employee work hard and efficiently. But the workers incentive might be to expend as minimal an effort as possible but still perform their assigned duties. If the employer cannot detect that the employee is not giving maximum effort, the employee is free to give a diminished effort. In these situations, it would be unsurprising for employees to give less than full effort.

The key to these situations of misaligned incentives is the presence of hidden actions: efforts (or lack thereof) on the part of one or both parties that are unobserved by the other. For example, if a job simply requires and agent putting two component pieces of a computer together on an assembly line, the principal can presumable observe the agents’ actions through monitoring the number of parts they assemble in an hour. They can also test to make sure the connection of each set of two parts is good. In this case, it is easy to write a contract to align incentives: specify a number of parts the agent is required to join properly in an hour. But consider the situation where the principal cannot observe the effort of the agent. For example, in retail sales it might be difficult to observe each interaction with customers. We call these hidden actions and they make it difficult or impossible to write a contract specifying a particular level of effort.

Principal-agent relationships where there is hidden action can be addressed through contracts that seek to align incentives through rewarding performance instead of effort. A typical one in the case of retail sales is through the use of commission where the agent gets a percentage of every sale they make (or a percentage of sales over a certain threshold). This helps align the incentives of the principal, to make as much revenue as possible, with the agent who, with a commission contract, would like to maximize the value of their own sales. Other examples are CEO pay where the CEO of a company is compensated partially based on the performance of the firm itself.

23.4 Moral Hazard

Learning Objective 23.4: Explain moral hazard and how it can affect the efficiency of markets.

Another consequence of hidden action is Moral Hazard where people who have entered into contract to mitigate the cost of risk engage in riskier behavior because the costs have diminished. An insurance contract that covers the cost of damage from an automobile collision might cause the holder of the contract to drive in a less cautious manner, increasing the risk of an accident. Borrowers who have limited liability contracts might be more willing to take risks with the money they borrow. If the actions of the contract holder could be monitored, the problem would go away as the contract could be written to take behavior into account and either prohibit it or to charge a price based on the nature of the actions the contract holder engages in. It is the hidden action aspect of these contracts that gives rise to moral hazard.

The consequence of contracts that reduce the cost of risk taking and therefore increasing the incentives to takes risks is that they become more expensive. An insurer who sells auto insurance will have to charge higher premiums as a consequence of moral hazard because its costs go up with riskier driving on the part of the policy holders. This causes premiums to go up for all holders, making it too expensive for the most careful drivers and causing a market failure.

SUMMARY

Review: Topics and Related Learning Outcomes

23.1 The Market for Lemons Problem

Learning Objective 23.1: Describe the lemons problem in markets with asymmetric information.

23.2 Adverse Selection

Learning Objective 23.2: Explain the term adverse selection and how it affects insurance markets.

23.3 Principal-Agent Models

Learning Objective 23.3: Describe how asymmetric information can affect principal-agent relationships.

23.4 Moral Hazard

Learning Objective 23.4: Explain moral hazard and how it can affect the efficiency of markets.

Learn: Key Terms and Graphs

Terms

Hidden Characteristics

Hidden Action

Adverse Selection

Moral Hazard

Principal-Agent Relationships