| paper | d4-journals-jet-2021-0010 |
| authors | — |
| venue | ? |
| filed under | unclassified |
| judged by | gpt-6-luna / xhigh (triple__luna__xhigh__c6r1) |
| judge confidence | high |
| authors would recognise it | yes |
Theorem 3 establishes existence, but gives no algorithm, complexity bound, or computational formulation used to compute an allocation. The proposed type-level search is a plausible follow-on, yet it cannot supply the missing computational character to a result of this paper. The paper therefore fails bit (a).
fails bit a — no named computational result to mirror
The proposed formulation adapts Theorem 3's constrained fairness and efficiency conditions to type-level assignments; it does not cover the paper's other existence results or its equilibrium-support theorem.
There is a plausible continuous-population version of this paper’s allocation problem, especially for controlled school choice. But the paper has no named computational result to anchor it to under your criteria. Theorem 1, Theorem 2, Theorem 3, Corollary 1, and Theorem 4 establish existence or equilibrium support; none classifies an algorithmic problem as tractable or hard. The closest result is Theorem 3, proved here: for concave utilities and feasible constraints, for every \(\varepsilon>0\) there is an allocation that is \(\varepsilon\)-individually rational, \(\varepsilon\)-Pareto optimal, and has no equal-type \(\varepsilon\)-justified envy; there is also an allocation that is individually rational, weakly Pareto optimal, and has no strong equal-type justified envy. The unnumbered remark about computational hardness with coarse priorities is not an eligible anchor, and concerns a different assignment problem.
The strongest positive scenario is a large school district using lotteries and controlled composition rules. Let each type \(t\) specify a student’s complete utility profile, demand limit, reservation utility, and constraint class. Let \(\mu_t\) be that type’s population share, with only \(\tau\) supported types in a much larger population. Normalize school capacities and composition quotas per student. The variable \(x^t\) is the lottery or bundle assigned to each member of type \(t\); feasibility includes \(\sum_t\mu_t x^t_l=q_l\) for each school \(l\), along with the normalized quotas. This changes the population into the continuous object; the fact that assignments are lotteries is already part of the paper’s outcome model. Repeated preference and entitlement profiles are plausible in a district with standardized preference reports and a small set of protected or neighborhood categories, though the support-size assumption needs evidence for a real district.
If existence results are accepted as nearby, non-computational anchors, the lead candidate is the search problem Continuum Equal-Type NJE Allocation. Its input is finite rational type data \((\mu_t,v_t,c_t,r_t,\kappa(t))\), where \(v_t\) is a linear utility vector, \(c_t\) a demand limit, \(r_t\) a reservation utility, and \(\kappa(t)\) the class of agents interchangeable under the allocation constraints, plus an explicit rational polytope of feasible type-level assignments and a promise that an individually rational feasible assignment exists. Its output is an assignment \(x\) in that polytope such that every type is individually rational; there is no feasible \(y\) with \(v_t\cdot y^t>v_t\cdot x^t\) for every type; and, for every \(t,s\) with \(\kappa(t)=\kappa(s)\), it is not the case that both \(v_t\cdot x^s>v_t\cdot x^t\) and \(v_s\cdot x^t>r_s\). The last condition is exactly the paper’s no-strong-equal-type-justified-envy condition, expressed for the type-level assignment. This is a symmetric allocation question: each member of a type receives the same assignment.
The authors would likely recognize this as their problem: it preserves participation constraints, Pareto efficiency, justified envy, and controlled-allocation constraints, and Theorem 3 explicitly treats fairness among agents distinguished by constraints. As a conjecture for this rational linear search version, I would lean toward tractability: the type-level resource constraints form a finite-dimensional polytope, and the population count no longer drives the representation. But Theorem 3 is an existence theorem, not an algorithm; the pairwise envy conditions introduce disjunctions, and the paper gives no basis for a polynomial-time claim. The further questions are whether the search is polynomial in \(L,\tau\), and the input bit length, and whether its solutions approximate or otherwise characterize the corresponding finite high-multiplicity allocation problem.
The weakest point is substantial: this candidate is not anchored by a qualifying computational result, and the type-level symmetric formulation needs justification against the paper’s full allocation space. Also, the exact utility and reservation profiles must repeat often enough for \(\tau\) to stay moderate. So this is a credible population mirror of Theorem 3’s *allocation question*, but the paper itself does not establish the computational mirror that the programme is looking for.
The strongest case against is that the paper’s named results are existence and equilibrium-support theorems, not computational results. Theorem 3 is the proponent’s best anchor, but it proves existence under concavity and convex feasibility; it gives no algorithm, finite input representation, or complexity claim. The proposed search problem adds all three by restricting utilities to rational linear data, requiring an explicitly represented polytope, and asking for an algorithm to find an allocation. That could be a worthwhile new project, but it is computational theory built on an existence theorem—not a continuous mirror of a computational result in this paper. Under ChoCo’s stated gate, that is the decisive objection.
The school-choice setting itself is credible. Repeated standardized preference and entitlement profiles could produce a large population with moderate support, and the theorem’s equal-type fairness condition is a natural anchor for an allocation question. High multiplicity supports that scenario; it is not a novelty collision. Nor does the fact that the paper already allows lotteries dispose of a population mirror. Those points make the proposed setting sensible, but they do not supply the missing computational result.
There is also a modeling wrinkle in carrying Theorem 3’s fairness condition into a nonatomic population. A single student has zero mass, so swapping two students’ assignments leaves aggregate quotas unchanged. The paper’s constraint-aware notion turns on whether a pairwise swap is feasible; in the limit, that feasibility test may stop distinguishing constraint classes. One can retain equal-type justified envy as the fairness criterion—the authors would recognize it—but then it is an imposed type-level condition, not a consequence of the continuum’s pairwise-swap feasibility. Resolving that distinction would take additional modeling work.
So I would reject this as a *continuous computational mirror of the paper*, while conceding that computing Theorem 3’s allocations in high-multiplicity school-choice instances could be a worthwhile follow-on question. The negative case is strongest on scope, not on the claim that no useful algorithmic problem could be developed from the paper.
The adversarial triple: the proponent anchors on up to three named results; the opponent sees that case and must defeat every anchor; the judge decides which case convinced it. These are the pipeline’s own outputs, generated by tools/triple_run.py — no human edited them. The paper’s own text is not reproduced here beyond the quoted statement above.