Mechanism Design

A mechanism is incentive-compatible when every participant's dominant strategy is to truthfully reveal their private information. This property eliminates strategic misrepresentation, ensuring the mechanism designer can trust the inputs received.

• Dominant Strategy Incentive Compatibility (DSIC): Truth-telling is always the best response regardless of what others do.
• Bayesian Incentive Compatibility (BIC): Truth-telling is optimal only when others are also truthful, assuming common prior beliefs.
• The classic Vickrey-Clarke-Groves (VCG) mechanism achieves DSIC for allocating public goods by making each participant's payment equal to the externality they impose on others.

A mechanism satisfies individual rationality when no participant is made worse off by joining the game than by abstaining. This participation constraint is essential for voluntary systems.

• Ex-post IR: A participant never regrets joining, regardless of the final outcome.
• Interim IR: Participation is beneficial in expectation, given the participant's private information.
• Without IR, agents would simply refuse to engage, causing the mechanism to collapse. This property is critical in designing supply chain auctions where suppliers must be incentivized to bid rather than bypass the platform.

A mechanism is budget-balanced when the total payments collected from some participants exactly equal the total disbursements to others, with no external subsidy required. This property ensures financial self-sufficiency.

• Strong Budget Balance: The sum of all transfers is exactly zero.
• Weak Budget Balance: The mechanism may run a surplus but never a deficit.
• In practice, achieving simultaneous incentive compatibility, individual rationality, and budget balance is often impossible due to the Myerson-Satterthwaite impossibility theorem, forcing designers to prioritize which properties are non-negotiable for their application.

A mechanism achieves allocative efficiency when resources are assigned to the participants who value them most highly, maximizing total social welfare. This is the canonical objective in market design.

• In a double auction, allocative efficiency means every unit is transferred from a seller with a lower valuation to a buyer with a higher valuation.
• The VCG mechanism is allocatively efficient because it aligns each agent's payment with the social opportunity cost of their allocation.
• In supply chain contexts, allocative efficiency translates to matching scarce production capacity to the highest-value customer orders, a direct application of prescriptive analytics.

A mechanism is strategy-proof if truth-telling is a weakly dominant strategy for every participant—meaning no alternative strategy ever yields a strictly better outcome, regardless of others' actions.

• Strategy-proofness is a stronger condition than Nash incentive compatibility; it requires optimality against all possible opponent strategies, not just equilibrium ones.
• The Gibbard-Satterthwaite theorem proves that no non-dictatorial voting mechanism with three or more outcomes can be strategy-proof, establishing a fundamental limit.
• Practical mechanisms often relax this to strategy-proofness in the large, where truth-telling becomes optimal as the number of participants grows, enabling tractable market designs.

The revelation principle states that for any equilibrium of any arbitrary mechanism, there exists an equivalent direct revelation mechanism where agents truthfully report their types and the outcome is identical.

• This foundational theorem dramatically simplifies mechanism design: instead of searching over all possible complex games, the designer can restrict attention to direct mechanisms where the only action is reporting a type.
• The principle assumes the designer can commit to the outcome function and that agents have quasi-linear utilities.
• In practice, this enables the design of combinatorial auctions for logistics procurement, where carriers simply submit their true cost functions and the system computes optimal allocations.

The foundational theorem of mechanism design states that any outcome achievable by an arbitrary mechanism can also be achieved by a direct-revelation mechanism where truth-telling is a Bayesian Nash equilibrium. In supply chains, this translates to designing procurement auctions where suppliers' dominant strategy is to bid their true marginal cost rather than shading bids strategically.

• Vickrey-Clarke-Groves (VCG) auctions achieve this by charging each winner the externality they impose on others, not their own bid
• In logistics, a Groves mechanism can elicit truthful capacity reports from carriers by making payment a function of reported capacity's impact on system-wide cost
• The principle eliminates the bullwhip effect's information distortion at its source: strategic misreporting

In a warehouse with autonomous mobile robots (AMRs) or a port with automated guided vehicles (AGVs), a central planner must assign tasks to agents with private information about their battery state, current location, and maintenance needs. A strategy-proof mechanism ensures no agent's onboard planner has an incentive to misreport its state to avoid undesirable assignments.

• The Serial Dictatorship mechanism: agents are ordered randomly, and each picks its most preferred task from what remains—simple and strategy-proof but not Pareto-efficient with transfers
• Competitive equilibrium from equal incomes (CEEI): each agent receives an equal budget of artificial currency and a market-clearing price vector emerges, yielding envy-free, Pareto-efficient allocations
• These mechanisms prevent the tragedy of the commons where individual agents hoard easy tasks and starve the system of overall throughput

A fundamental tension in mechanism design is the Myerson-Satterthwaite impossibility theorem: no mechanism can simultaneously achieve ex-post efficiency, individual rationality, incentive compatibility, and budget balance in bilateral trade with private values. In supply chain coordination, this forces explicit trade-offs.

• A VCG mechanism achieves efficiency and incentive compatibility but runs a deficit—the auctioneer must inject external funds because payments collected are less than the total value transferred
• In practice, the deficit is funded by membership fees or considered an acceptable cost for truthful information revelation
• Alternative: dAGVA mechanisms (expected externality mechanisms) achieve budget balance but only Bayesian incentive compatibility—truth-telling is optimal only in expectation, not for every possible realization of others' types

When a supply disruption occurs—a port closure, a supplier bankruptcy, a weather event—the mechanism must reallocate capacity in real time. Dynamic pivot mechanisms extend VCG to sequential settings where agents' private information evolves and actions have intertemporal consequences.

• At each decision epoch, the marginal contribution of each agent is computed as the difference between optimal system value with and without that agent's participation
• The pivot payment equals the externality imposed: agent i pays the total value others would have received if i were absent, minus what others actually receive given i's presence
• This creates a Markov perfect equilibrium where truthful reporting of evolving private state (e.g., remaining capacity after a disruption) is optimal at every time step, enabling autonomous re-optimization without human negotiation