Designing Incentive-Compatible Contracts

The Futility of Naive Questions

Let's revisit the problem of adverse selection with a specific scenario. Imagine a couple, Hannah and Charles, are considering a dynamic with a new partner, Ben. A critical point of negotiation is health and safety protocols. Ben has private information about his past behavior and his resulting comfort level with risk, which we can simplify into two "types": Low-Risk (prefers to always use protection) and High-Risk (is comfortable forgoing protection with tested partners).

If Hannah simply asks, "Are you Low-Risk or High-Risk?", what should she expect? A High-Risk Ben, wanting to make a good impression and not be seen as reckless, has a strong incentive to misrepresent himself as Low-Risk. This strategic incentive makes direct questions unreliable for revealing private information.

This is a universal problem in settings with private information. To see a classic illustration of this, please watch the following video from Nobel laureate Eric Maskin. He discusses the challenge faced by a government trying to sell radio spectrum licenses to the company that values them most, without knowing their true valuations.

As Maskin explains, both naive questioning and a seemingly straightforward first-price auction create incentives for bidders to be dishonest—either by exaggerating or understating their true value. The design of the "game" itself encourages strategic misrepresentation. The challenge, then, is to design a better game.

A Smarter Game: The Power of Incentives

The solution, as Maskin brilliantly illustrates, is not to demand truth, but to make it the most profitable strategy. He introduces the second-price auction, a mechanism that elegantly solves the problem.

The second-price auction is a beautiful example of an incentive-compatible mechanism. It works because the rules are structured such that your optimal strategy is to be honest, regardless of what others do. You cannot gain from lying, and you might lose.

This brings us to a foundational insight: if we can design a set of rules where partners are naturally led to reveal their true preferences, we can overcome the problem of adverse selection.

The Revelation Principle

The genius of the second-price auction is a specific case of a much more general theorem: the Revelation Principle.

In simple terms, the Revelation Principle states that for almost any complex, indirect mechanism you can imagine (e.g., a multi-stage negotiation, a series of costly signals, probationary periods) that produces a certain outcome, there exists a much simpler direct mechanism that achieves the very same outcome, in which everyone is incentivized to truthfully report their private information.

This principle is a cornerstone of mechanism design because it radically simplifies the designer's task. Instead of inventing elaborate games, we can restrict our attention to designing simple, direct contracts and focus on getting the incentives right. To do that, we need to satisfy two key constraints.

The following lecture notes provide a formal treatment of these ideas. Given your background in Bayesian decision theory, the mathematical formulation should feel familiar.

To complement these formal definitions, the following video walks through a very clear, practical example of calculating these constraints.

So, to design a successful "contract" that reveals private information, it must satisfy:

• Individual Rationality (IR): Each partner, whatever their "type," must find the proposed agreement better than their outside option (e.g., walking away).
• Incentive Compatibility (IC): Each partner, whatever their "type," must get a better outcome by telling the truth than by lying and pretending to be another type.

Designing a Relationship Contract

Let's apply this to our scenario with Hannah, Charles, and Ben. Their goal is to design a direct mechanism to determine the sexual health protocol based on Ben's true risk preference.

• Principal: Hannah and Charles
• Agent: Ben
• Agent's Types: Low-Risk-Tolerance or High-Risk-Tolerance
• Mechanism: A direct "contract" where Ben declares his type.
• Outcomes (The Social Choice Function):
  • If Ben declares Low-Risk: The outcome is Protocol-C (Comfort). All parties agree to use barriers for all relevant sexual acts.
  • If Ben declares High-Risk: The outcome is Protocol-T (Trust). The parties agree to be barrier-free, but only after all three provide verified, negative STI test results from within the past 14 days.

Now, let's analyze the incentives, just as was done in the video. We must ensure the contract is both IR and IC.

Individual Rationality (IR):

• A Low-Risk Ben must prefer Protocol-C to walking away. (This is likely true, as it matches his preference).
• A High-Risk Ben must prefer Protocol-T to walking away. (This depends on whether the pleasure of being barrier-free outweighs the hassle of getting tested).

Incentive Compatibility (IC):
The crucial part is preventing beneficial lies.

1. Would a Low-Risk Ben ever lie and declare High-Risk? No. That would lead to Protocol-T, exposing him to a level of risk he is uncomfortable with. His utility for Protocol-C is higher than for Protocol-T. This side of the constraint is satisfied.
2. Would a High-Risk Ben ever lie and declare Low-Risk? Maybe. This would lead to Protocol-C. He might do this if the pleasure of being barrier-free (in Protocol-T) is less than the cost/hassle of getting tested.

The job of the mechanism designer (Hannah and Charles) is to ensure this second condition doesn't happen. They must design Protocol-T to be sufficiently attractive to the High-Risk type.

Let $U_{High}(P_T)$ be the utility a High-Risk type gets from Protocol-T, and $U_{High}(P_C)$ be his utility from Protocol-C. The IC constraint is:

To satisfy this, Hannah and Charles can manipulate the utilities by adjusting the "costs" and "benefits" within the protocols. For example, they could lower the "cost" of Protocol-T by offering to help schedule the lab appointments or covering the testing fees. This increases $U_{High}(P_T)$, making it more likely that a High-Risk Ben will truthfully choose the option intended for him.

This flowchart shows the key properties a mechanism must satisfy. Our focus today has been on designing for Incentive Compatibility (IC) and Individual Rationality (IR) to solve the problem of hidden information.

Information Revelation in the Real World

This is not just a theoretical exercise. Interventions that make it easier for people to signal their private information have been shown to have dramatic real-world effects. A study of the marriage market in Malawi provides a compelling case.

In this study, an "opt-out" HIV testing program acted as a mechanism. It lowered the cost and stigma of testing, making it easy for "safe" individuals to reveal their type. The result? The adverse selection problem was reduced, and "safe" people were able to match and marry faster because the information asymmetry was resolved.

Conclusion

In this lesson, we've unlocked a central tool for navigating negotiations under asymmetric information. Instead of trying to guess at a partner's true preferences, we can design the rules of the game itself to make honesty the most logical and beneficial choice.

Key Takeaways:

• Revelation Principle: A foundational concept stating that if a desirable outcome is achievable at all, it's achievable with a simple direct mechanism where everyone is incentivized to be truthful.
• Direct Mechanism: A "contract" where participants' only action is to report their private information or "type."
• Incentive Compatibility (IC): The condition that truth-telling must yield a higher (or equal) payoff than lying for every type of participant.
• Individual Rationality (IR): The condition that every participant must find the outcome of participating better than their outside option.
• Design, Don't Guess: By carefully crafting the outcomes associated with each reported type, we can create a system that is self-enforcing and encourages honesty.

Preview of the Next Lesson:

Now that you understand the theory of direct, incentive-compatible mechanisms, we will put it into practice. In the next lesson, we will "Draft a direct mechanism (contract) to solve a specific adverse selection problem, such as revealing true STI testing history." We will move from theory to application, creating a detailed, practical agreement from scratch.