Near-feasible Fair Allocations in Two-sided Markets

Andr\'es Moraga; Javier Cembrano; Victor Verdugo

REVIEW 3 major objections 3 minor 2 cited by

Near-feasible Fair Allocations in Two-sided Markets

T0 review · 3 major / 3 minor · reviewed 2026-08-07 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read A single rounding theorem turns any fractional allocation into an integer one with tunable fairness, capacity, and demand deviations.

desk verdict The core rounding theorem is a real contribution, but two of the three headline applications—Theorem 2 and Theorem 5—have parameter gaps that mean the stated results do not follow from the main theorem as written. read the letter →

arxiv 2506.01178 v1 pith:IYQAZKPH submitted 2025-06-01 cs.GT econ.THmath.OC

Javier Cembrano , Andrés Moraga , Victor Verdugo This is my paper

classification cs.GTecon.THmath.OC MSC 91B3290C2791B68

keywords iterativeroundingnear-feasibleallocationstwo-sidedmarketsgroupfairnessmultidimensionalapportionmentenvy-freenessstablematchingwithcouplesresourceallocation

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper is trying to establish a general recipe for near-feasible fair allocations in two-sided markets: start with any fractional allocation, then round it so that each group's utility, each resource's usage, and the total demand deviate only by amounts the policy designer chooses in advance. Its main theorem gives one sufficient condition under which this rounding always succeeds in polynomial time, and the condition makes the trade-off explicit, allowing more fairness deviation to buy smaller capacity deviation and vice versa. If true, this would give a single toolkit for problems that are usually treated separately, including school assignment, stable matching with couples, and political apportionment, each with fairness constraints layered on top of feasibility. A sympathetic reader would care because the theorem converts infeasible fairness requirements into feasible approximate ones with controllable slack.

What carries the argument

The load-bearing object is the iterative rounding algorithm together with the constraint-counting inequality (7), $\frac{\mathbf{1}_{\psi=1}}{2} + \sum_{\ell\in[d]} \frac{1}{\alpha_\ell+1} + \frac{\omega^*}{\delta+1} \le 1$. The algorithm works on a hypergraph whose hyperedges are agent-bundle pairs and whose vertices are agents, resources, and groups; each iteration solves the linear program LP(I,u,·) with equality constraints for agents whose fractional allocation sums to one, for groups with at least $\alpha_\ell+1$ fractional entries, for resources with at least $\delta+1$ fractional units, and optionally for total demand. The inequality (7) ensures this linear program always has more variables than linearly independent equality constraints, or that the equality constraints are dependent, so an extreme point makes progress; the same counting controls the final deviations from the starting fractional allocation.

What would settle it

Check the $d=3$ apportionment example with $\alpha=(2,2,2)$: the claimed house-size deviation $\Delta=2$ requires some integer $\delta$ that satisfies the strict lower-bound case of Theorem 1 while the per-resource and house-size deviations coincide in the single-resource construction; if exhaustive search shows no such $\delta$ exists, then Theorem 5's improvement over Cembrano et al. [2022] does not follow.

Watch

Extended reading notes

Core claim

The central discovery is a rounding theorem for the multidimensional capacitated resource allocation (MCRA) model, where agents are partitioned into groups along several dimensions, each agent demands bundles of resources, and each resource has a capacity. Theorem 1 says that for any fractional allocation, any utility functions, and any deviation parameters $\alpha$, $\delta$, $\Delta$ satisfying the condition $\frac{\mathbf{1}_{\psi=1}}{2} + \sum_{\ell\in[d]} \frac{1}{\alpha_\ell+1} + \frac{\omega^*}{\delta+1} \le 1$, with $\Delta$ either at least 2 or large enough relative to the slack, there is an integer rounding whose group utilities deviate by less than $\alpha_\ell$ times the maximum utility in the group, whose per-resource usage deviates by less than $\delta$, and whose total demand deviates by less than $\omega^*$ times $\Delta$, computable in polynomial time. The proof runs an iterative rounding algorithm that solves a linear program at each step, freezing groups and resources with many fractional entries; the condition guarantees the system always has more variables than linearly independent equality constraints, so an extreme point strictly reduces fractionality. The paper then derives consequences for school allocation, stable matching with couples, and multidimensional apportionment, obtaining strengthened near-feasibility guarantees in each.

Load-bearing premise

The load-bearing premise is that the rounding theorem still applies when no separate bound on each resource's overuse is given, because in the apportionment construction the per-resource deviation and the house-size deviation are the same number; the claimed small house-size deviation depends on there being a concrete such bound that satisfies the theorem's condition.

Editorial extensions

If this is right

For any MCRA instance and any initial fractional allocation, a polynomial-time rounding exists with tunable deviations, so policymakers can choose $\alpha$, $\delta$, and $\Delta$ to match the application's priorities.
For school assignment with concave group-fairness objectives, the framework gives approximately fair allocations with constant trade-offs such as $(\alpha,\delta)=(2,4)$, $(3,2)$, and $(5,1)$ when demands are unit-sized.
The new ratio-based group envy-freeness notion admits approximately envy-free allocations whose deviation depends on the number of groups per dimension rather than the number of agents, breaking the linear-in-$|A|$ impossibility.
For stable matching with couples, the framework yields stable, approximately fair allocations with small capacity deviations, recovering the known $\delta=\Delta=2$ bounds and adding fairness constraints.
For multidimensional apportionment, the framework bounds the total deviation from the house size, not just per-dimension deviations, whenever the dimension-wise slack condition holds.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

A direct test of the theorem's reach would encode refugee resettlement or healthcare rationing as MCRA instances, since their capacity and fairness constraints fit the same three-parameter deviation structure.
The ratio-based group envy-freeness notion could be probed beyond group-homogeneous instances; if a greedy fractional construction still works, the rounding argument would extend to heterogeneous utilities.
For a fixed market, the sufficient condition defines an explicit Pareto frontier of admissible $(\alpha, \delta, \Delta)$ triples; enumerating that frontier on real school-choice data would turn the existential guarantee into a policy menu.

Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

The MCRA framework and the iterative rounding theorem are genuinely useful. At the center is a Beck-Fiala style result with a tuned sufficient condition that trades off group utility deviations, per-resource capacity violations, and total demand deviation. That Theorem 1 is worth having, and the proof structure—the LP with more variables than constraints, the termination argument, the deviation bounds for groups, resources, and total demand—looks sound to me. The new ratio-based relaxed envy-freeness notion in Theorem 3 is also a good idea, and the claim that it breaks the linear-in-|A| impossibility from Procaccia et al. is plausible if the appendix proof holds up. The paper is also honest about where it is thin, notably the stable-matching part where it relies on an unknown-to-be-polynomial subroutine.

But the application sections have real gaps, and they are not typos. Theorem 2 applies Theorem 1 with Δ=0. That is outside the theorem's hypotheses: condition (i) requires Δ≥2, and condition (ii) requires a positive lower bound plus a strict inequality. Even worse, inequality (6) with Δ=0 demands |Σ ω_a(y−x)| < 0, which is impossible. So the school-allocation result as proven is empty. This is a load-bearing problem, not a stylistic complaint.

The stress-test note on Theorem 5 is also correct. The proof of the apportionment bound never specifies the per-resource deviation δ. For the example α=(2,2,2), condition (7) becomes 3/4 + 1/(δ+1) ≤ 1. If δ=3, the inequality is an equality, so condition (ii) is unavailable, and condition (i) requires ψ=1, which contradicts the proof's ψ=0 choice. If δ≥4, condition (ii) forces Δ ≥ 1/(1−3/4−1/(δ+1))−1, which for δ=4 is already 19, far above the claimed bound of 2. So the claimed improvement over Cembrano et al. does not follow from the proof as written.

I want to stress that I think these are patchable. The main theorem is structurally solid, and the intended applications are credible. But they need actual mathematical work, not a footnote: Theorem 2 needs either a different choice of Δ or a modified argument, and Theorem 5 needs either a specific δ that works or a revised statement if the house-size bound is not attainable through this route. The appendix envy-free fractional allocation proof is also compressed and needs more careful treatment of the continuous limit.

This paper deserves a serious referee. The framework and Theorem 1 are interesting enough on their own, and the gaps are specific and diagnosable, which is exactly what referees are for. I would send it out, with the request that the referee focus on the Theorem 2 and Theorem 5 derivations.

Referee Report

3 major / 3 minor

Summary. The paper introduces MCRA, a multidimensional two-sided resource-allocation model with group structure, and proves a general iterative-rounding theorem (Theorem 1) that produces integer allocations close to a given fractional allocation in three respects: group utility deviations, per-resource capacity deviations, and total demand deviation. The authors then instantiate this framework in three market-design settings: school allocation with group fairness and a new relaxed envy-freeness notion (Theorems 2 and 3), stable matching with couples (Theorem 4), and multidimensional political apportionment (Theorem 5), where they claim an improved house-size deviation bound over Cembrano, Correa, and Verdugo (2022). The main rounding theorem has a coherent proof structure based on a rank-counting lemma and a linear-programming extreme-point argument. The applications vary in maturity: Theorem 4 fits the framework cleanly, while the proofs of Theorems 2 and 5 contain gaps in their invocation of Theorem 1.

Significance. If fully established, the paper would provide a flexible, unified rounding tool for a broad class of fairness-constrained allocation problems, with concrete constant-deviation guarantees in school choice, stable matching with couples, and apportionment. The framework is well-motivated and the parameterized trade-off between utility loss, per-resource capacity violation, and total excess is a useful contribution. The paper also contains a constructive proof of the main theorem and explicitly connects to prior impossibility and near-feasibility results, which is valuable. However, the current proof gaps in two of the three flagship applications materially weaken the advertised claims, in particular the claimed improvement over Cembrano et al. in multidimensional apportionment.

major comments (3)

[§4.3, Theorem 5 and its proof] The proof of Theorem 5 applies Theorem 1 with ψ=0, α'=α+1, and ω*=1, but never specifies the per-resource deviation parameter δ required by condition (7). For the paper's own highlighted example α=(2,2,2), we have Σ 1/(αℓ+2) = 3/4. If δ=3, then (7) holds with equality, so condition (ii) of Theorem 1 is unavailable and condition (i) would require ψ=1, which makes the left-hand side of (7) equal to 1/2+3/4+1/4 > 1. If δ≥4, then (7) holds strictly but condition (ii) imposes Δ ≥ 1/(1−3/4−1/(δ+1)) − 1 ≥ 19, so Theorem 1 cannot deliver the strict bound <3 on total deviation that the proof uses to conclude the claimed house-size deviation Δ=2. Thus the asserted improvement over Cembrano et al. (2022) is not established by the proof as written.
[§4.1, Theorem 2 and its proof] The proof invokes Theorem 1 with ψ=1, δ+1, and 'Δ=0'. This is outside the hypotheses of Theorem 1: condition (i) requires Δ≥2 when ψ=1, and condition (ii) is unavailable because the displayed inequality in Theorem 2 is non-strict and also gives a positive lower bound on Δ. Since the subsequent capacity-excess argument uses only the per-resource deviation bound (31) and does not use the total-deviation bound from Theorem 1, replacing Δ=0 by Δ=2 would likely repair the application; nevertheless, as written the invocation is formally invalid.
[§3, Theorem 1, condition (ii)] The theorem's condition (ii) explicitly requires that (7) hold strictly before the displayed lower bound on Δ applies. In Lemma 1's proof, the subcase with χ_t=1 and Δ satisfying (ii) correctly uses this strictness, but in Theorem 2 and Theorem 5 the applications do not verify strictness in the way the theorem demands. The two preceding comments are instances of this more general issue; the statement of Theorem 1 itself appears internally consistent.

minor comments (3)

[§4.3, proof of Theorem 5] The notation for the bound ⌈1/(1−Σ1/(αℓ+2))−2⌉ in the theorem statement and the value ⌈1/(1−Σ1/(αℓ+2))−1⌉ used inside the proof is confusing: the latter is the Δ parameter passed to Theorem 1, while the former is the final integer house-size deviation. Please distinguish these two values explicitly.
[§4.1, proof of Theorem 2, final displayed inequality] In the last line of the proof, the expression '(ω−1)|A|' appears to be a typo for '(ω∗−1)|A|'; the preceding line uses ω∗ consistently.
[§4.3, definitions] When defining the MCRA instance in the proof of Theorem 5, the agents are A=E×[c] and the groups are inherited from the MA instance; it would help to write explicitly how each group G_{ℓ,i} in the MCRA instance is defined, since the proof currently relies on this identification without stating it.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: Theorem 1 is self-contained and the cited prior results are independent external theorems.

full rationale

The central derivation is self-contained. Theorem 1 is proved directly from Algorithm 1, Lemmas 1 and 2, and a variable/constraint counting argument over the linear program; the sufficient condition (7) is derived from the analysis rather than assumed as the conclusion. The applications (Theorems 2, 4, and 5) invoke Theorem 1 as a black box with explicit parameter choices, and the deviation bounds are consequences of the theorem rather than fitted inputs relabeled as predictions. The main self-citations are to Cembrano et al. (2022) and Nguyen and Vohra (2018), but these are used as external mathematical facts: optimal solutions of [LP-MA] are exactly proportional apportionments, and any rounding of a dominating extreme point is stable. Those theorems have independent proofs and do not incorporate the house-size bound or fairness deviations that the present paper claims, so they are not circular. No equation in the paper reduces to its own input by construction, and no fitted value is renamed as a prediction. Separate from circularity, I note a soundness concern in the proof of Theorem 5: the resource-deviation parameter δ is never specified when Theorem 1 is applied, and for α=(2,2,2) no (δ,Δ) pair simultaneously satisfies condition (7) and the (i)/(ii) alternatives while yielding the claimed house-size bound Δ=2. That is a correctness gap in the derivation as written, not a circular step.

Assumptions & free parameters 0 free parameters · 4 assumptions · 0 invented entities

The paper introduces a mathematical framework and no new physical or economic entities. The parameters α, δ, Δ are policy choices, not fitted values. The main assumptions are fractional feasibility and cited characterization theorems from prior work.

assumptions (4)

domain assumption Fractional feasibility of the assignment-MCRA instance for Theorems 2 and 4.
Theorems 2 and 4 require the instance to admit a fractional resource allocation; without fractional feasibility there is no starting point for the rounding procedure.
domain assumption Nguyen and Vohra [2018, Lemma 3]: any rounding of a dominating extreme point of [LP-Stable] is stable.
Theorem 4 inherits stability from this cited result, and the paper itself notes that finding a dominating extreme point is not known to be polynomial-time solvable.
domain assumption Cembrano et al. [2022, Theorem 1]: optimal solutions of [LP-MA] characterize proportional apportionments.
Theorem 5 relies on this characterization to ensure proportionality after rounding.
domain assumption The continuous limit of Algorithm 2 yields a fractional allocation satisfying the group envy-freeness inequality.
Appendix A gives a one-paragraph proof of Claim 5 via the eating-algorithm continuous limit, but the saturation-time argument is not fully formalized.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Near-feasible Fair Allocations in Two-sided Markets." pith.science (2026). https://pith.science/paper/IYQAZKPH

@misc{pith2026250601178,
  author       = {Pith},
  title        = {Pith review of: Near-feasible Fair Allocations in Two-sided Markets},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/IYQAZKPH}},
  note         = {Machine review of arXiv:2506.01178}
}

read the original abstract

We study resource allocation in two-sided markets from a fundamental perspective and introduce a general modeling and algorithmic framework to effectively incorporate the complex and multidimensional aspects of fairness. Our main technical contribution is to show the existence of a range of near-feasible resource allocations parameterized in different model primitives to give flexibility when balancing the different policymaking requirements, allowing policy designers to fix these values according to the specific application. To construct our near-feasible allocations, we start from a fractional resource allocation and perform an iterative rounding procedure to get an integer allocation. We show a simple yet flexible and strong sufficient condition for the target feasibility deviations to guarantee that the rounding procedure succeeds, exhibiting the underlying trade-offs between market capacities, agents' demand, and fairness. To showcase our framework's modeling and algorithmic capabilities, we consider three prominent market design problems: school allocation, stable matching with couples, and political apportionment. In each of them, we obtain strengthened guarantees on the existence of near-feasible allocations capturing the corresponding fairness notions, such as proportionality, envy-freeness, and stability.

Discussion (0). Sign in to comment.

Forward citations

Cited by 2 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

Designing Pairwise-Stable Agent Seating Arrangements
cs.DS 2026-07 accept novelty 6.0 of 10

Designable target graphs plus stable-partition bundles yield poly-time pairwise-stable seating, team, and b-matching arrangements, with hardness when the graph is given.
Equity by Design? On the Trade-Offs in Fairness-Driven Recommendation in Heterogeneous Two-Sided Markets
cs.GT 2026-02 conditional novelty 6.0 of 10

The 'free fairness' result for producer constraints vanishes for multi-item recommendations; a CVaR group-fairness objective and business constraints can be added with moderate trade-offs.

Reference graph

Works this paper leans on

43 extracted references · 41 canonical work pages · cited by 2 Pith papers

[1]

S. K. A., K. Munagala, M. Nasre, G. S. Sankar, et al. Approximation algorithms for school assignment: Group fairness and multi-criteria optimization. arXiv preprint arXiv:2403.15623, 2024

work page Pith review arXiv 2024
[2]

Ahani, T

N. Ahani, T. Andersson, A. Martinello, A. Teytelboym, and A. C. Trapp. Placement optimization in refugee resettlement. Operations Research, 69 0 (5): 0 1468--1486, 2021

work page 2021
[3]

Andersson and L

T. Andersson and L. Ehlers. Assigning refugees to landlords in sweden: Efficient, stable, and maximum matchings. The Scandinavian Journal of Economics, 122 0 (3): 0 937--965, 2020

work page 2020
[4]

Arnosti, C

N. Arnosti, C. Bonet, and J. Sethuraman. Explainable affirmative action. In Proceedings of the 25th ACM Conference on Economics and Computation, pages 310--310, 2024

work page 2024
[5]

Aziz and F

H. Aziz and F. Brandl. Efficient, fair, and incentive-compatible healthcare rationing. In Proceedings of the 22nd ACM Conference on Economics and Computation, pages 103--104, 2021

work page 2021
[6]

Balinski and G

M. Balinski and G. Demange. An axiomatic approach to proportionality between matrices. Mathematics of Operations Research, 14 0 (4): 0 700--719, 1989 a

work page 1989
[7]

Balinski and G

M. Balinski and G. Demange. Algorithms for proportional matrices in reals and integers. Mathematical Programming, 45 0 (1-3): 0 193--210, 1989 b

work page 1989
[8]

Balinski and H

M. Balinski and H. P. Young. Fair representation: Meeting the ideal of one man, one vote. Brookings Institution Press, 2010

work page 2010

Show all 43 references

[9]

N. Bansal. Constructive algorithms for discrepancy minimization. In Proceedings of the 51st Annual Symposium on Foundations of Computer Science, pages 3--10, 2010

2010
[10]

Bansal and S

N. Bansal and S. Garg. Algorithmic discrepancy beyond partial coloring. In Proceedings of the 49th Annual ACM SIGACT Symposium on Theory of Computing, pages 914--926, 2017

2017
[11]

Bansal, D

N. Bansal, D. Dadush, and S. Garg. An algorithm for koml \'o s conjecture matching banaszczyk's bound. SIAM Journal on Computing, 48 0 (2): 0 534--553, 2019

2019
[12]

Bansal, H

N. Bansal, H. Jiang, S. Singla, and M. Sinha. Online vector balancing and geometric discrepancy. In Proceedings of the 52nd Annual ACM SIGACT Symposium on Theory of Computing, pages 1139--1152, 2020

2020
[13]

Beck and T

J. Beck and T. Fiala. Integer-making theorems. Discrete Applied Mathematics, 3 0 (1): 0 1--8, 1981

1981
[14]

S. Bera, D. Chakrabarty, N. Flores, and M. Negahbani. Fair algorithms for clustering. Advances in Neural Information Processing Systems, 32, 2019

2019
[15]

Bobbio, M

F. Bobbio, M. Carvalho, A. Lodi, I. Rios, and A. Torrico. Capacity planning in stable matching: An application to school choice. In Proceedings of the 24th ACM Conference on Economics and Computation, pages 295--295, 2023

2023
[16]

Bubeck et al

S. Bubeck et al. Convex optimization: Algorithms and complexity. Foundations and Trends in Machine Learning, 8 0 (3-4): 0 231--357, 2015

2015
[17]

B. Bukh. An improvement of the beck--fiala theorem. Combinatorics, Probability and Computing, 25 0 (3): 0 380--398, 2016

2016
[18]

Cembrano, J

J. Cembrano, J. Correa, and V. Verdugo. Multidimensional political apportionment. Proceedings of the National Academy of Sciences, 119 0 (15): 0 e2109305119, 2022

2022
[19]

Cembrano, J

J. Cembrano, J. Correa, G. D \' az, and V. Verdugo. Proportionality in multiple dimensions to design electoral systems. arXiv preprint arXiv:2410.03304, 2024

2024 arXiv
[20]

Chandrasekaran, Y

K. Chandrasekaran, Y. Faenza, C. He, and J. Sethuraman. Scarf's algorithm on arborescence hypergraphs. To appear in ICALP, 2025

2025
[21]

Delacr \'e taz, S

D. Delacr \'e taz, S. D. Kominers, and A. Teytelboym. Refugee resettlement. University of Oxford Department of Economics Working Paper, 2016

2016
[22]

Faenza, C

Y. Faenza, C. He, and J. Sethuraman. Scarf's algorithm and stable marriages. Mathematics of Operations Research, 2025

2025
[23]

Feldman, J

M. Feldman, J. Garg, V. V. Narayan, and T. Ponitka. Proportionally fair makespan approximation. In Proceedings of the AAAI Conference on Artificial Intelligence, pages 13839--13846, 2025

2025
[24]

Gaffke and F

N. Gaffke and F. Pukelsheim. Divisor methods for proportional representation systems: An optimization approach to vector and matrix apportionment problems. Mathematical Social Sciences, 56 0 (2): 0 166--184, 2008

2008
[25]

Gale and L

D. Gale and L. S. Shapley. College admissions and the stability of marriage. The American Mathematical Monthly, 69 0 (1): 0 9--15, 1962

1962
[26]

Haeringer

G. Haeringer. Market design: auctions and matching. MIT Press, 2018

2018
[27]

Immorlica, F

N. Immorlica, F. Echenique, and V. V. Vazirani. Online and Matching-based Market Design. Cambridge University Press, 2023

2023
[28]

Korte and J

B. Korte and J. Vygen. Combinatorial optimization, volume 1. Springer, 2011

2011
[29]

L. C. Lau, R. Ravi, and M. Singh. Iterative methods in combinatorial optimization, volume 46. Cambridge University Press, 2011

2011
[30]

Lovett and R

S. Lovett and R. Meka. Constructive discrepancy minimization by walking on the edges. SIAM Journal on Computing, 44 0 (5): 0 1573--1582, 2015

2015
[31]

Makarychev and A

Y. Makarychev and A. Vakilian. Approximation algorithms for socially fair clustering. In Conference on Learning Theory, pages 3246--3264, 2021

2021
[32]

Mathieu and V

C. Mathieu and V. Verdugo. Apportionment with parity constraints. Mathematical Programming, 203 0 (1): 0 135--168, 2024

2024
[33]

H. Moulin. Fair division and collective welfare. MIT press, 2004

2004
[34]

Nguyen, T

H. Nguyen, T. Nguyen, and A. Teytelboym. Stability in matching markets with complex constraints. Management Science, 67 0 (12): 0 7438--7454, 2021

2021
[35]

Nguyen and R

T. Nguyen and R. Vohra. Near-feasible stable matchings with couples. American Economic Review, 108 0 (11): 0 3154–69, 2018

2018
[36]

Nguyen and R

T. Nguyen and R. Vohra. Stable matching with proportionality constraints. Operations Research, 67 0 (6): 0 1503--1519, 2019

2019
[37]

o nmez, M. U. \

P. A. Pathak, T. S \"o nmez, M. U. \"U nver, and M. B. Yenmez. Fair allocation of vaccines, ventilators and antiviral treatments: leaving no ethical value behind in health care rationing. In Proceedings of the 22nd ACM Conference on Economics and Computation, pages 785--786, 2021

2021
[38]

Procaccia, I

A. Procaccia, I. Robinson, and J. Tucker-Foltz. School redistricting: Wiping unfairness off the map. In Proceedings of the 2024 Annual ACM-SIAM Symposium on Discrete Algorithms, pages 2704--2724, 2024

2024
[39]

Pukelsheim

F. Pukelsheim. Proportional Representation. Springer International Publishing, 2017

2017
[40]

A. E. Roth. Marketplaces, markets, and market design. American Economic Review, 108 0 (7): 0 1609--1658, 2018

2018
[41]

Rothvoss

T. Rothvoss. Constructive discrepancy minimization for convex sets. SIAM Journal on Computing, 46 0 (1): 0 224--234, 2017

2017
[42]

Saha and A

B. Saha and A. Srinivasan. A new approximation technique for resource-allocation problems. Random Structures & Algorithms, 52 0 (4): 0 680--715, 2018

2018
[43]

H. P. Young. Equity: in theory and practice. Princeton University Press, 2020

2020

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.

[1] [1]

S. K. A., K. Munagala, M. Nasre, G. S. Sankar, et al. Approximation algorithms for school assignment: Group fairness and multi-criteria optimization. arXiv preprint arXiv:2403.15623, 2024

work page Pith review arXiv 2024

[2] [2]

Ahani, T

N. Ahani, T. Andersson, A. Martinello, A. Teytelboym, and A. C. Trapp. Placement optimization in refugee resettlement. Operations Research, 69 0 (5): 0 1468--1486, 2021

work page 2021

[3] [3]

Andersson and L

T. Andersson and L. Ehlers. Assigning refugees to landlords in sweden: Efficient, stable, and maximum matchings. The Scandinavian Journal of Economics, 122 0 (3): 0 937--965, 2020

work page 2020

[4] [4]

Arnosti, C

N. Arnosti, C. Bonet, and J. Sethuraman. Explainable affirmative action. In Proceedings of the 25th ACM Conference on Economics and Computation, pages 310--310, 2024

work page 2024

[5] [5]

Aziz and F

H. Aziz and F. Brandl. Efficient, fair, and incentive-compatible healthcare rationing. In Proceedings of the 22nd ACM Conference on Economics and Computation, pages 103--104, 2021

work page 2021

[6] [6]

Balinski and G

M. Balinski and G. Demange. An axiomatic approach to proportionality between matrices. Mathematics of Operations Research, 14 0 (4): 0 700--719, 1989 a

work page 1989

[7] [7]

Balinski and G

M. Balinski and G. Demange. Algorithms for proportional matrices in reals and integers. Mathematical Programming, 45 0 (1-3): 0 193--210, 1989 b

work page 1989

[8] [8]

Balinski and H

M. Balinski and H. P. Young. Fair representation: Meeting the ideal of one man, one vote. Brookings Institution Press, 2010

work page 2010