Wasserstein Stability and Free Boundaries in Measure-Parameterized Bilevel Obstacle Problems
Kun Huang
Competent, clean, honestly-scoped theoretical paper whose sharp exponent result and policy/boundary distinction are quotable, but whose components are standard and whose impact is confined to a narrow subfield.
Abstract
We study obstacle-constrained variational problems whose reduced energies depend on a probability law through a lower-level optimizer. Uniform strong convexity yields a single-valued Lipschitz follower response, while convexity and a Poincare inequality give a unique upper-level policy. A type-Lipschitz reduced marginal then implies Lipschitz continuity of the policy map from the 1-Wasserstein metric to the energy space. In a one-dimensional linear-obstacle subclass, the policy derivative is the positive part of a cumulative forcing. Single crossing makes the coincidence set an interval. An algebraic crossing of order m gives a modulus for its endpoint; odd-power examples show that this exponent is sharp, while a transversal crossing recovers Lipschitz stability. For empirical laws on compact subsets of , dimension-dependent Wasserstein bounds yield finite-sample rates for policies and thresholds. At a transversal population root, the empirical threshold is asymptotically linear, with an explicit influence function and central limit theorem. For higher-dimensional regular patches, a conditional level-set argument gives local Hausdorff stability when a nondegenerate switching function is available. An explicit quadratic firm response produces a nonlinear corporate-tax schedule with an endogenous zero-tax region and a Wasserstein-stable threshold. The tax illustration is analytic and uses no empirical calibration.
AI Impact Assessments
(1 models)Impact Assessment
Core Contribution. This single-author theoretical paper studies obstacle-constrained variational problems whose forcing depends on a probability law through a lower-level optimizer, and tracks a full perturbation chain: population law → follower response → obstacle policy → free boundary. The central conceptual message is a clean separation between two distinct stability notions: stability of the *optimizer* (the policy map, controlled by strong monotonicity + Poincaré, yielding Wasserstein-Lipschitz continuity in energy norm) versus stability of the *active-set boundary* (which requires an additional nondegenerate "switching function"). The most distinctive technical result is that an algebraic crossing of order *m* in the cumulative forcing yields a modulus for the free-boundary point, with an explicit odd-power (cubic) family demonstrating that the exponent is sharp. The paper further transfers deterministic moduli to empirical laws (finite-sample rates via DKW and Fournier–Guillin/Weed–Bach bounds) and derives a CLT with explicit influence function at transversal roots. A stylized corporate-tax model provides a fully closed-form illustration.
Methodological Rigor. The proofs are sound, self-contained, and carefully staged through an explicit assumption hierarchy (Table 1). Each result rests on well-established machinery—strong convexity for the follower, standard VI existence/uniqueness, the positive-part contraction for the 1D policy, coupling estimates for the Wasserstein bounds. The sharpness example is genuinely convincing and the cubic family is embedded in the measure framework via point masses, closing off the alternative that the exponent could be improved. The author is admirably honest that the components ("scalar root perturbation, empirical-Wasserstein, central-limit ingredients") are "standard in isolation" and that the contribution is their integration. The higher-dimensional result (Theorem 8.3) is explicitly *conditional*—it assumes rather than derives a stable nondegenerate defining function, and the author correctly frames it as a "reusable final step" rather than new geometric theory. This intellectual honesty is a strength but also delimits the depth of the contribution.
Potential Impact. The likely influence is modest and confined to a small community at the intersection of PDE-constrained optimization, stochastic programming, and variational inequality sensitivity analysis. The sharp exponent result and the policy/boundary distinction are the ideas most likely to be cited or reused. The corporate-tax application is explicitly stylized ("analytic verification case," "uses no empirical calibration") and deliberately omits general equilibrium, incidence, and incentive-compatibility—so it will not directly influence public-finance practice. The paper does not improve the deeper obstacle-problem literature it cites (Figalli–Serra, Serfaty–Serra, Christof–Wachsmuth), and openly says so.
Timeliness & Relevance. The framing rides two active currents—optimal-transport metrics and bilevel optimization—which gives it surface relevance. However, the specific problem class (measure-parameterized bilevel obstacle problems with a scalar switching function) is narrow, and the sharp results essentially require the one-dimensional linear-obstacle subclass. It addresses a coherent gap rather than a pressing bottleneck.
Strengths.
Limitations.
Other observations. Reproducibility of the mathematics is high—proofs are complete and a small script reproduces the single illustrative figure. The work requires solid graduate-level familiarity with variational inequalities, functional analysis, and optimal transport, but no specialist frontier expertise. Resource intensity is trivial (pure theory). The paper does not contest or corroborate any prior empirical claim. Overall this is a competent, well-crafted, honest theoretical paper whose primary value is a clean conceptual distinction and one sharp modulus result, with a likely narrow but real readership.
Generated Sep 9, 2026
Comparison History (0)
Competent, clean, honestly-scoped theoretical paper whose sharp exponent result and policy/boundary distinction are quotable, but whose components are standard and whose impact is confined to a narrow subfield.