Game Theory and Partnership Agreements

The Foundation: The Revelation Principle

Before we evaluate different agreement structures, it's useful to know a fundamental "good news" theorem of mechanism design. In essence, it tells us that if we're trying to design an agreement to achieve a certain goal (like efficiency or fairness), we don't need to consider infinitely complex, multi-stage communication protocols. We can restrict our attention to simple, direct mechanisms where everyone is asked to state their private information (their "type" or preferences), and it's in their best interest to do so honestly. This is the Revelation Principle.

The Revelation Principle is our starting point. It assures us that designing agreements that promote honesty is a feasible and central part of creating stable, efficient relationships. However, it doesn't guarantee that any arbitrary agreement structure will have this property. Most real-world "agreements" are not explicitly designed mechanisms and can create powerful incentives against truthful revelation.

When Relationship Structure Dictates Strategy: A Case Study

To understand how a relationship's structure can distort preference revelation, we'll examine a fascinating study that, while focused on polygamous households in Senegal, provides a direct and powerful analogy for the strategic dynamics within a polyandrous vee.

In this context, you can think of the two co-wives as the two non-hinge partners (P1, P2) and the shared husband as the hinge (H). The study models the wives' decisions about fertility rates as a non-cooperative game. A wife's fertility choice is a costly, visible action that reveals her "demand" for the husband's resources, both current and future (e.g., through inheritance for her children).

Let's unpack the strategic tension revealed by this model. The "strategic response" term, $B_i=\frac{{ϵ}_i-{\theta }_{hi}}{1+{\theta }_{hi}+{\theta }_{ni}}$, shows how a wife $i$'s optimal number of children $n_i^{\ast }$ depends on the number of children her co-wife has, $n_{-i}$.

• The term ${ϵ}_i$ represents co-wife rivalry. This is an "overbidding" effect. The more children the other wife has, the more you want, to compete for the husband's resources and status.
• The term ${\theta }_{hi}$ represents the weight given to the husband's preferences. Since the husband has a target for his total number of children, more children for the other wife means he desires fewer from you. This is a "substitution" effect, where you reduce your "demand" to align with the hinge's overall goal.

The sign of $B_i$ determines whether children are strategic complements ($B_i>0$, rivalry dominates) or substitutes ($B_i<0$, aligning with the husband dominates). The wife's "revealed preference" (her fertility choice) is not a simple statement of her own desires. It is a strategic calculation that balances rivalry against the desire to please the husband.

Translating to the Vee:

This provides a sharp analytical lens for your goal. The "agreement structure" is the vee itself. A non-hinge partner's communication of needs (for time, attention, emotional support) is their preference revelation.

• Overbidding Effect: If P1 perceives that P2 is getting a lot of the hinge's time, P1 might strategically inflate their own stated needs to compete. They might ask for more time than they truly need, anticipating that this is necessary to secure even a satisfactory amount. This is not truthful revelation; it's strategic inflation driven by the competitive structure.
• Substitution Effect: If the hinge (H) expresses that they are feeling overwhelmed and desires overall harmony and less demand on their time, P1 might strategically downplay their needs. They might pretend to be more low-maintenance than they really are to align with the hinge's stated goal and appear to be the "easier" partner. This is also not truthful revelation; it's strategic suppression.

The key insight is that the very existence of the vee structure forces partners into a meta-game where they must decide how to represent their preferences. Truthful revelation becomes just one strategy among many, and often not the optimal one. The paper's empirical finding—that a wife's fertility decisions change based on the characteristics of her co-wife—is compelling evidence that this strategic calculus is real and powerful.

The Shadow of the Future: Can the Agreement Be Changed?

So far, we have assumed that the "rules of the game" are fixed. But in any real relationship, agreements are fluid and can be renegotiated. This possibility casts a long shadow on present behavior, further complicating the decision to reveal preferences truthfully.

A well-designed mechanism often relies on credible threats for off-equilibrium behavior (i.e., punishments for lying). For example, a rule might state, "If you are found to have misrepresented your needs, you forfeit your planned vacation with the hinge." The VCG payment is another form of this. However, if all parties know that such a harsh outcome would be so unpleasant for everyone that they would simply renegotiate it away, the threat is not credible.

The possibility of renegotiation acts as a powerful reality check on any proposed relationship agreement. If an agreement relies on punishments that are not "renegotiation-proof," it will fail to guide behavior effectively.

Consider a vee agreement for time allocation. Suppose P1 lies, claiming to have a critical need for the hinge's help over a weekend that was promised to P2. The lie is discovered. The pre-agreed punishment is that P1 gets no time with H for a month. This is a very painful outcome for both P1 and H. It is highly likely they, along with P2, would renegotiate this to a less severe punishment. Knowing this in advance, the initial deterrent against lying is weakened. The agreement structure is not robust, and it fails to promote truthful revelation.

Conclusion

In this lesson, we have moved from the idealized world of VCG to the more complex reality of evaluating how different agreement structures influence truthful preference revelation.

• Key Takeaways:
  • The Revelation Principle gives us a theoretical foundation, showing that if any outcome is achievable, it's achievable with a simple, direct, and truthful agreement. The challenge lies in finding that agreement.
  • The very structure of the relationship creates strategic incentives that can distort preference revelation. As seen in the polygamy case study, partners may engage in "overbidding" (due to rivalry) or "substitution" (to appease the hinge), moving their revealed preferences away from their true ones.
  • Agreements must be renegotiation-proof. The possibility of changing the rules later can undermine the credibility of an agreement's enforcement mechanisms, making it easier for partners to deviate from truthful behavior.
  • Evaluating an agreement structure requires analyzing not just its explicit rules, but also the implicit strategic game it creates and whether its outcomes are stable against future renegotiation.

This concludes our module on designing rules and agreements. We have explored how to model vee relationships, analyze their strategic dynamics, and design mechanisms to promote desirable outcomes like efficiency and honesty.

In our next module, we will shift our analytical lens. We will move from the non-cooperative framework of managing individual incentives to the domain of Cooperative Games and Coalition Formation. We will start by asking not how to constrain self-interest, but how partners can form cooperative coalitions and how the value generated by their cooperation can be divided fairly.