| paper | Veto Core Consistent Preference Aggregation |
| authors | — |
| venue | AAMAS 2024 |
| filed under | voting · theory |
| judged by | gpt-5.6-luna / xhigh (triple__luna__xhigh__c2r1) |
| judge confidence | high |
| authors would recognise it | yes |
The paper contains no numbered theorem, lemma, corollary, or proposition asserting a computational complexity or algorithmic result, so bit (a) fails. The continuous veto-by-consumption model is direct and author-recognisable, but it restates Definition 3 and cannot compensate for the missing computational anchor.
fails bit a — no named computational result to mirror
The mirror covers veto-by-consumption evaluation and its high-multiplicity interpretation, but leaves the paper’s numbered axiomatic results without qualifying computational counterparts.
The strongest honest positive answer is qualified: this paper has an exceptionally natural continuous-population mirror, but it contains no qualifying named computational anchor.
Every numbered result is axiomatic. For example, Proposition 16 proves that veto by consumption is strictly monotonic, and Proposition 18 proves positive responsiveness after generalized-antiplurality tie-breaking. Neither asserts that a computational problem is in \(P\), NP-hard, FPT, or similar. The paper’s computational statements are unnumbered, while polynomial-time computation of the proportional veto core is attributed to [23]. Thus, strictly under the requested anchoring rule, there are zero eligible anchors.
The best positive mirror is nevertheless very concrete. The paper’s veto-by-consumption rule already treats the population continuously: replace the number of voters of each ranking type by its mass.
Take a finite candidate set \(C\), the type set \(T\) of rankings of \(C\), and a rational distribution \(\mu\in\mathbb{Q}_{\ge 0}^{T}\) with \(\sum_t\mu_t=1\). Every candidate has capacity \(1\). At any stage with surviving candidates \(A\subseteq C\), each type sends all its mass to its least-preferred candidate in \(A\). Thus candidate \(c\) is eaten at rate
\[ \rho_c(A,\mu)=\sum_{t:\,c\text{ is least preferred in }A\text{ under }t}\mu_t. \]
If \(q_c\) is the remaining capacity, the next phase lasts
\[ \Delta=\min_{c:\rho_c(A,\mu)>0}\frac{q_c}{\rho_c(A,\mu)}. \]
Update \(q_c\leftarrow q_c-\Delta\rho_c(A,\mu)\), remove every candidate whose capacity reaches zero, and reassign its eating mass to the next-worst surviving candidate. The winner set \(W_{\mathrm{VbC}}(\mu)\) consists of the candidates removed in the final phase.
This gives the precise problem:
\[ \textsf{VbC-Winner}_{\infty} \]
Given \((C,\mu)\), compute \(W_{\mathrm{VbC}}(\mu)\), or decide whether a designated candidate \(c^\star\) belongs to it. A solution is the exact winner set, represented by rational arithmetic. There is no welfare objective: the decision variable is the induced mass-consumption schedule, and the rule declares the last surviving candidate or candidates winners.
This is a genuine population continuization, not continuity of the outcome space. The outcome remains a discrete candidate; what becomes continuous is the electorate and the rate at which each ranking type consumes candidates. If \(\mu_t=n_t/n\), then the discrete profile with \(n_t\) voters of type \(t\) has exactly the same elimination order: all eating rates are merely multiplied by \(n\). Hence the bridge to high multiplicity is exact.
The regime is plausible for the paper’s own applications: large public consultations over a fixed set of windfarm sites, nuclear-waste policies, or other multi-criteria alternatives, with many participants belonging to recurring stakeholder, geographical, or institutional categories. Each category supplies a complete ranking type, and \(n\gg\tau\), where \(\tau\) is the number of distinct rankings actually represented. The paper’s authors would recognise this mirror immediately, since Definition 3 explicitly describes veto by consumption as a continuous version of voting by veto.
If a structural proposition is accepted as a near-anchor, my lead is Proposition 16, proved in this paper with the proof placed in the supplementary material: “Veto by consumption is strictly monotonic.”
Its continuous counterpart is:
\[ \textsf{VbC-Promotion}_{\infty} \]
Given \((C,\mu)\), a candidate \(c\), a type \(t\), and a type \(t'\) obtained from \(t\) by moving \(c\) upward by one position, together with \(0<\delta\le\mu_t\), define
\[ \mu'=\mu-\delta e_t+\delta e_{t'}. \]
Compute \(W_{\mathrm{VbC}}(\mu')\), and decide whether
\[ c\in W_{\mathrm{VbC}}(\mu') \qquad\text{and}\qquad W_{\mathrm{VbC}}(\mu')\subseteq W_{\mathrm{VbC}}(\mu). \]
For rational \(\delta\), Proposition 16 lifts directly by scaling to a discrete profile and repeatedly applying the one-voter statement. Thus the expected classification is Class A: exact winner computation is already polynomial in the number of types and candidates, and the monotonicity conclusion follows from the paper’s structural theorem. This is a faithful continuous form of the paper’s result, although not itself a new complexity theorem.
A second computational object naturally suggested by the paper is continuous proportional-veto-core membership. For candidate \(c\), let
\[ \alpha_B(c,\mu)= \sum_{t:\,b\succ_t c\text{ for every }b\in B}\mu_t. \]
The natural continuous core condition is
\[ c\in\mathrm{VCore}_{\infty}(\mu) \quad\Longleftrightarrow\quad \alpha_B(c,\mu)\le 1-\frac{|B|}{m} \quad\text{for every }B\subseteq C\setminus\{c\}. \]
The strict reverse inequality means that a coalition with enough mass can block \(c\). The paper cites [23] for polynomial computation of the proportional veto core, so this is supporting evidence for a tractable Class A mirror, though again it is not a numbered computational result of this paper.
The genuinely new ChoCo questions would be control and robustness around these rules. For example, given conversion costs \(\gamma(t,t')\), one could ask for the minimum-cost mass transfer \(x_{t,t'}\) such that
\[ \mu'_u=\mu_u-\sum_{t'}x_{u,t'}+\sum_t x_{t,u} \]
makes \(c^\star\) a veto-by-consumption winner, a unique winner under generalized-antiplurality tie-breaking, or a member of the continuous veto core. That is a proper continuous bribery/control problem with a meaningful objective
\[ \min_x \sum_{t,t'}\gamma(t,t')x_{t,t'}. \]
The core version looks promising for Class A because the blocking conditions are linear in \(\mu'\) and the cited core algorithm supplies a separation strategy. The VbC winner version may be more difficult because the elimination order changes across piecewise-linear regions; it could yield either a tractable parametric optimization problem or a continuum-specific hardness result.
My weakest point is decisive: the paper’s contribution is axiomatic, not computational. Moreover, its basic continuous rule is already present in Definition 3 and earlier work, so one cannot present \(\textsf{VbC-Winner}_{\infty}\) as a new computational discovery. The positive case is therefore that the paper supplies an unusually well-motivated, author-recognisable continuous population model from which new computational questions can be generated—not that the paper itself contains a qualifying ChoCo complexity result.
On ChoCo’s stated standard, this paper is a poor source paper for continuization. Its numbered results are entirely axiomatic: Propositions 6, 8, 13–30 and the open problems concern consistency, monotonicity, participation, and independence. None gives a complexity classification, algorithmic theorem, or computational hardness result.
The proposed \(\textsf{VbC-Winner}_{\infty}\) is not a new mirror. It is exactly Definition 3 rewritten with masses. If \(\mu_t=n_t/n\), every consumption rate is the corresponding discrete rate divided by \(n\), so the elimination order is unchanged and all phase durations are merely rescaled. That establishes a sensible high-multiplicity regime, but the paper already presents veto by consumption as a continuous rule and states that it is computable with polynomially many arithmetic operations. The proposed problem is therefore evaluation of an already-defined, already-computable rule, not a missing computational question exposed by the paper.
The same defeats Proposition 16. Its continuous version is either the vacuous statement that changing one voter has no effect, or the finite-mass statement obtained by moving \(\delta\) mass from one type to another. For rational \(\delta\), that follows by replicating voters and repeatedly applying the proposition. The repaired statement is mathematically legitimate, but it is an axiom, not a computational problem. Asking for the minimum mass or minimum cost needed to create a winner would be a new control problem invented around veto by consumption; the paper supplies neither the intervention model nor a theorem that makes this particular control problem distinctive. The proponent’s displayed decision test also omits the necessary premise that \(c\) was a winner before the promotion.
Continuous proportional-veto-core membership is the strongest rescue, but it is not an anchor from this paper. The paper explicitly attributes polynomial-time core computation to Ianovski and Kondratev [23]. The mass formulation
\[ \sum_{t:\,b\succ_t c\ \forall b\in B}\mu_t \le 1-\frac{|B|}{m} \]
is the natural high-multiplicity encoding of that existing object. A bit-complexity analysis for rational \(\mu\), or minimum-cost mass transfer into the core, might be worthwhile as ChoCo research. But it would be a continuation of the prior computational literature, not a continuous mirror of one of this paper’s results. The control variant would likewise be attachable to almost any anonymous voting rule.
The negative case should not claim that rankings lack multiplicity. They plainly admit a convincing regime: large consultations on windfarm sites, waste policies, or other finite agendas with recurring stakeholder types. Nor should it claim that the continuous rule is ill-defined or trivial. The proponent is right on both points.
The decisive objection is narrower: the paper’s central contribution is axiomatic, while its only genuinely computational objects—veto-by-consumption evaluation and proportional-veto-core computation—are already continuous or inherited from earlier work. What remains is a natural baseline plus generic control questions. If ChoCo accepts any new optimization problem built around any voting rule, the universal negative is weak; continuous core-control is a plausible project. Under the programme’s stated requirement to mirror computational results of the paper, however, I would reject this paper as having no worthwhile continuous mirror.
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.