Complexity of Strong Implementability

We consider the question of implementability of a social choice function in a classical setting where the preferences of finitely many selfish individuals with private information have to be aggregated towards a social choice. This is one of the central questions in mechanism design. If the concept of weak implementation is considered, the Revelation Principle states that one can restrict attention to truthful implementations and direct revelation mechanisms, which implies that implementability of a social choice function is easy to check. For the concept of strong implementation, however, the Revelation Principle becomes invalid, and the complexity of deciding whether a given social choice function is strongly implementable has been open so far. In this paper, we show by using methods from polyhedral theory that strong implementability of a social choice function can be decided in polynomial space and that each of the payments needed for strong implementation can always be chosen to be of polynomial encoding length. Moreover, we show that strong implementability of a social choice function involving only a single selfish individual can be decided in polynomial time via linear programming.

💡 Research Summary

The paper tackles a long‑standing open problem in mechanism design: determining the computational complexity of strong implementability of a social choice function. A social choice function aggregates the private preferences of finitely many self‑interested agents into a collective decision. Under the weak implementation paradigm, the Revelation Principle guarantees that any implementable outcome can be achieved by a truthful direct‑revelation mechanism, making the verification of implementability essentially a matter of checking incentive‑compatibility constraints, which can be done in polynomial time. Strong implementation, however, requires that the desired outcome be the result of every Nash equilibrium of the mechanism, not just some equilibrium. In this setting the Revelation Principle breaks down, and no efficient general method for testing strong implementability had been known.

The authors first formalize the strong implementation problem as a system of linear inequalities that capture the requirement that, for every possible type profile and every equilibrium strategy profile, the mechanism’s outcome coincides with the prescribed social choice. This system defines a high‑dimensional polyhedron in the space of payment vectors. By invoking results from polyhedral theory, they show that the polyhedron is non‑empty if and only if there exists a payment scheme that strongly implements the function. Crucially, they prove that the existence test can be performed within polynomial space (PSPACE). Their algorithm does not need to enumerate all equilibria; instead it works with the implicit description of the polyhedron and uses space‑efficient procedures to decide feasibility.

A second major contribution is the bound on the size of the payments. The authors demonstrate that if a strongly implementing payment vector exists, there is always one whose components can be represented with a number of bits polynomial in the size of the input (i.e., the description of the agents’ type spaces, preferences, and the social choice function). This guarantees that the payments are not only theoretically existent but also practically encodable.

The paper then isolates the special case of a single selfish agent. With only one participant, strategic interaction disappears, and the strong implementation condition collapses to a set of linear constraints ensuring that the agent’s optimal response always yields the desired outcome. The authors show that this reduced problem can be expressed as a linear program and solved in polynomial time, placing the single‑agent strong implementation problem in the class P.

Overall, the work establishes a clear complexity landscape: strong implementability is PSPACE‑complete in the general multi‑agent setting, yet tractable (polynomial‑time) when the number of agents is one. The authors discuss the implications for real‑world mechanism design, such as public‑good allocation or complex auction formats, where strong implementation may be required to guarantee robustness against multiple equilibria. They also outline future research directions, including the development of approximation algorithms for large‑scale instances, parameterized complexity analyses based on the number of types or agents, and experimental validation of the proposed polyhedral approach. This contribution significantly advances our understanding of the algorithmic limits of robust mechanism design.