Adverse Selection vs. Moral Hazard

The Fundamental Distinction: Information vs. Action

Games of incomplete information create what is often called a Principal-Agent problem. In this framework, a Principal (e.g., an employer, an insurer) designs a contract for an Agent (e.g., an employee, an insured person) to perform a task. The information asymmetry gives rise to two distinct challenges, depending on when the hidden element comes into play.

A great way to start is with a direct and clear definition of each.

As the video explains:

• Adverse Selection is a problem of hidden information. The Agent has private information about their "type" before entering into a contract. The Principal doesn't know this information and risks contracting with an undesirable type.
• Moral Hazard is a problem of hidden action. The Agent takes an unobservable action after the contract is agreed upon. The Principal cannot verify if the Agent is behaving as desired.

The timing is the crucial difference. The following diagrams, from the resource Games and Information, illustrate this perfectly.

Let's break down each concept with classic examples.

Adverse Selection: The Problem of Hidden Information

Adverse selection occurs when one party in a negotiation has relevant information that the other party lacks. This pre-contractual asymmetry can lead to a "selection" of participants that is "adverse" to the uninformed party.

The Market for Lemons

The most famous example is George Akerlof's 1970 paper on the market for used cars, which he called "The Market for Lemons."

The core logic is as follows:

1. Sellers know the quality of their car (e.g., Peach or Lemon). Buyers do not.
2. A buyer, unable to distinguish between them, will offer a price based on the average quality of cars on the market.
3. This average price might be below the value of a high-quality car (a Peach).
4. Consequently, sellers of Peaches will refuse to sell, pulling their cars from the market.
5. As Peaches leave, the average quality of cars on the market drops, further lowering the price buyers are willing to pay, until only Lemons remain. The market for good cars collapses.

The Insurance Death Spiral

The same logic applies forcefully to health insurance. Healthy individuals are "Peaches" and unhealthy individuals are "Lemons" from the insurer's perspective.

Let's formalize the expected cost calculation from the video.

• Population: 90% healthy, 10% sickly.
• Expected Cost (Healthy): $0.10\cdot \$10,000+0.90\cdot \$0=\$1,000$
• Expected Cost (Sickly): $0.50\cdot \$10,000+0.50\cdot \$0=\$5,000$
• Pool Average Cost: $0.90\cdot \$1,000+0.10\cdot \$5,000=\$900+\$500=\$1,400$

If the insurer sets a premium just above this, say at $1,500, to make a profit:

• Sickly people see a great deal: their expected cost is $5,000, but they only pay $1,500. They will definitely buy.
• Healthy people face a bad deal: their expected cost is only $1,000, but the premium is $1,500. Only those who are very risk-averse might buy.

If, as in the video's example, only half the healthy people buy, the pool of customers changes. The new pool is not 90/10 but 45 healthy people and 10 sickly people. The insurer's expected cost per customer is now much higher, and they lose money. Raising the price further just chases more healthy people away, hence the "death spiral."

Moral Hazard: The Problem of Hidden Action

Moral hazard occurs when an agent's behavior changes after a contract is signed because they are insulated from risk. The action is "hidden" because the principal cannot perfectly monitor the agent's effort or choices.

Insurance and Risky Behavior

Insurance is the classic context for moral hazard.

Once you have theft insurance for your bike, you might not be as diligent about locking it up. Once you have health insurance, you might engage in riskier hobbies. The key is that the insurance contract changes your incentives.

A stark real-world example is in the financial sector.

The Cost of Moral Hazard

This change in behavior isn't just a curiosity; it creates real economic costs, or "welfare loss." When an agent doesn't bear the full cost of their actions, they may exert less effort than is socially optimal.

The video explains two main costs:

1. Lower Efficiency: The Agent (e.g., an unemployed worker receiving benefits) takes "too much" leisure because they don't face the full cost of not working. The socially optimal balance between work and leisure is distorted.
2. Increased Costs for Others: The costs of the Agent's behavior are borne by the Principal (e.g., taxpayers funding unemployment benefits). This can require higher taxes, which in turn distorts the incentives of others.

Summary Table

Here is a table to consolidate the key distinctions:

For a more formal, mathematical treatment of these Principal-Agent problems, which you might find interesting, you can refer to documents like the notes from Mishra, particularly sections 2 and 3, which set up and analyze models for adverse selection and moral hazard, respectively. You can find these at [PDF] 1 The principal-agent problems 2 Adverse selection problem.

Conclusion

You've now explored the two primary forms of market failure caused by information asymmetry. This distinction is not just academic; it dictates the solutions.

• To fight adverse selection, principals need ways to screen for types (e.g., medical exams for insurance, offering different contracts to let agents self-select).
• To fight moral hazard, principals need ways to monitor actions or align incentives (e.g., deductibles in insurance, performance-based pay).

Key Takeaways:

• Adverse Selection is a pre-contractual problem of hidden information about an agent's type.
• Moral Hazard is a post-contractual problem of hidden action taken by an agent.
• The classic example of adverse selection is the "market for lemons," where bad types can drive out good types.
• The classic example of moral hazard is when insurance leads to riskier behavior, as the agent no longer bears the full cost of their actions.

This lesson concludes our module on static games of incomplete information. In the next module, we'll dive into a major real-world application where these informational problems are paramount: auctions. We will begin by comparing the strategic implications of first-price and second-price sealed-bid auctions.