pith. sign in

arxiv: 2606.19278 · v1 · pith:3J2SZQZOnew · submitted 2026-06-17 · 🧮 math.OC

Sequential Fair Allocation and Routing in Nonprofit Operations

Haiqing Gao , Seyed M.R. Iravani , Sean R. Sinclair This is my paper

Pith reviewed 2026-06-26 19:48 UTC · model grok-4.3

classification 🧮 math.OC
keywords sequential allocationfair allocationrouting policythreshold structurecoefficient of variationmax-min fairnessnonprofit operationsdemand uncertainty
0
0 comments X

The pith

Fair resource allocation to uncertain demands uses thresholds and visits high-variability locations first.

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

The paper examines how a central planner should distribute a limited divisible resource fairly across locations with uncertain demands while also choosing the sequence in which to visit them. It establishes that any fixed visit order admits an equating property that forces the optimal allocation into a threshold form: low realized demands are met in full while higher demands are met proportionally. It further shows that, under stated conditions, the best visit order is by decreasing coefficient of variation of demand. These structural results support a simple heuristic and allow experiments that map different fairness objectives to different realized equity measures.

Core claim

For any fixed visitation order the optimal allocation satisfies an equating property that produces a threshold structure: demand at each location is fully satisfied when sufficiently low and otherwise met proportionally. The optimal routing policy visits locations in decreasing order of the coefficient of variation of demand under certain conditions. A heuristic called PPA-deCV that follows this ordering closely approximates the fairness-efficiency frontier of the jointly optimal policy.

What carries the argument

The equating property of optimal allocations for fixed order, which equates marginal fairness contributions and thereby generates the threshold allocation rule together with the decreasing-CV routing order.

Load-bearing premise

Demand distributions are known in advance and independent across locations.

What would settle it

A concrete three-location instance with independent known demand distributions in which the fairness achieved by an increasing-CV visit order exceeds that of the decreasing-CV order.

read the original abstract

We study a dynamic fair sequential allocation problem in which a central planner distributes a divisible resource across multiple locations under demand uncertainty. Motivated by applications such as humanitarian relief and food distribution, we incorporate routing decisions into the planner's problem and jointly optimize allocation and visitation order under two max-min fairness objectives, ex-post and forward. We first reveal an equating property of the optimal allocation for any fixed visitation order. This property is crucial to establish that the optimal allocation follows a threshold structure: at each location, demand is fully satisfied when sufficiently low and otherwise met proportionally. We then characterize the optimal routing policy and show that, under certain conditions, visiting locations in decreasing order of coefficient of variation (CV) of demand is optimal. Building on this insight, we propose a simple heuristic, PPA-deCV, which closely approximates the fairness-efficiency frontier of the jointly optimal policy. Next, through extensive numerical experiments, we compare multiple fairness objectives across different fairness metrics, demonstrating that improvements in surrogate max-min objectives do not necessarily translate into improvements in fairness metrics. Furthermore, we identify which objective is aligned with which fairness metric, providing practical guidance on objective selection to achieve fairness-efficiency trade-offs.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 3 minor

Summary. The paper examines a dynamic fair sequential allocation and routing problem for divisible resources under demand uncertainty, motivated by nonprofit applications like humanitarian relief. For any fixed visitation order, it establishes an equating property of optimal allocations and proves that the optimal policy has a threshold structure (full satisfaction below a demand threshold, proportional allocation above). It characterizes the optimal routing policy and shows that, under stated conditions, visiting locations in decreasing order of demand coefficient of variation (CV) is optimal. A heuristic PPA-deCV is proposed to approximate the jointly optimal fairness-efficiency frontier. Numerical experiments compare multiple max-min fairness objectives against various fairness metrics, revealing that gains in surrogate objectives do not always improve the metrics and identifying alignments between objectives and metrics.

Significance. If the derivations hold, the equating property and threshold structure provide clean, reusable insights for stochastic fair allocation problems with routing. The CV-based routing result and the PPA-deCV heuristic offer a practical way to navigate the fairness-efficiency trade-off without solving the full joint optimization. The numerical finding that surrogate objectives and realized fairness metrics can diverge supplies concrete guidance for practitioners choosing objectives in humanitarian operations. The work ships explicit structural theorems derived from standard independent-demand assumptions and reproducible numerical comparisons, both of which strengthen its contribution to the operations-research literature on fair sequential decisions.

major comments (2)
  1. [§4] §4 (Routing characterization): The statement that decreasing-CV order is optimal 'under certain conditions' is load-bearing for the routing claim, yet the precise conditions (e.g., on demand distributions, number of locations, or resource level) are not cross-referenced to the equating-property derivation; without an explicit theorem statement linking the two, it is unclear whether the numerical instances satisfy the conditions.
  2. [Numerical experiments] Numerical experiments (fairness-metric comparisons): The claim that 'improvements in surrogate max-min objectives do not necessarily translate into improvements in fairness metrics' is central to the practical-guidance contribution, but the manuscript reports only point estimates without error bars, replication counts, or statistical tests; this weakens the ability to assess whether observed divergences are robust or artifacts of particular demand realizations.
minor comments (3)
  1. [§2] Notation for the two max-min objectives (ex-post and forward) should be introduced once with a clear table or equation block rather than redefined inline in multiple sections.
  2. [Numerical experiments] Figure captions for the fairness-efficiency frontiers should state the number of demand scenarios or replications used to generate each curve.
  3. [§5] The definition of the PPA-deCV heuristic should include a short pseudocode block or explicit algorithmic steps so readers can reproduce the reported approximation quality.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the constructive comments and the recommendation of minor revision. We address the two major comments point by point below.

read point-by-point responses

  1. Referee: [§4] §4 (Routing characterization): The statement that decreasing-CV order is optimal 'under certain conditions' is load-bearing for the routing claim, yet the precise conditions (e.g., on demand distributions, number of locations, or resource level) are not cross-referenced to the equating-property derivation; without an explicit theorem statement linking the two, it is unclear whether the numerical instances satisfy the conditions.

    Authors: We agree that greater explicitness is warranted. The equating property is stated as Theorem 1 in §3 and is invoked to prove the threshold structure, which is then used in the proof of the routing result (Theorem 3 in §4) under the maintained assumptions of independent demands possessing finite second moments. In the revision we will (i) restate the precise conditions in the statement of Theorem 3, (ii) add an explicit forward reference to Theorem 1 inside that statement, and (iii) insert a short paragraph in the experimental setup (§5) confirming that all generated instances satisfy the same independence and moment conditions. revision: yes

  2. Referee: [Numerical experiments] Numerical experiments (fairness-metric comparisons): The claim that 'improvements in surrogate max-min objectives do not necessarily translate into improvements in fairness metrics' is central to the practical-guidance contribution, but the manuscript reports only point estimates without error bars, replication counts, or statistical tests; this weakens the ability to assess whether observed divergences are robust or artifacts of particular demand realizations.

    Authors: The observation is correct; the reported figures are single-realization point estimates. To address this we will augment the numerical section with 30 independent replications per parameter combination, report means accompanied by standard errors, and add a short robustness paragraph noting that the qualitative pattern of divergence between surrogate objectives and realized metrics persists across replications. These additions will be confined to the existing experimental design and will not change the paper's conclusions. revision: yes

Circularity Check

0 steps flagged

No significant circularity identified

full rationale

The paper's core results—the equating property, threshold allocation structure for fixed order, and CV-based routing optimality—are derived directly from the explicit model premises of known independent demand distributions and the two max-min fairness objectives. These steps use standard stochastic optimization techniques without reducing any claimed prediction or structural result to a fitted input, self-citation chain, or definitional equivalence. Numerical experiments compare objectives to metrics but introduce no load-bearing fitted parameters that force the reported outcomes. The derivation chain is self-contained.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

Only the abstract is available; no explicit free parameters, axioms, or invented entities are stated in the provided text.

pith-pipeline@v0.9.1-grok · 5738 in / 1153 out tokens · 20290 ms · 2026-06-26T19:48:01.053701+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

123 extracted references · 18 canonical work pages

  1. [1]

    arXiv preprint arXiv:2503.05276 , year=

    Constrained Reinforcement Learning for the Dynamic Inventory Routing Problem under Stochastic Supply and Demand , author=. arXiv preprint arXiv:2503.05276 , year=

    work page arXiv

  2. [2]

    All Things Considered , publisher=

    Godoy, Maria , year=. All Things Considered , publisher=

  3. [3]

    ACM SIGMETRICS Performance Evaluation Review , volume=

    Online fair allocation of perishable resources , author=. ACM SIGMETRICS Performance Evaluation Review , volume=. 2023 , publisher=

    2023

  4. [4]

    Coronavirus (Covid-19) Data in the United States , year =

  5. [5]

    and Shah, Arnav and Sah, Pratha and Moghadas, Seyed M

    Schneider, Eric C. and Shah, Arnav and Sah, Pratha and Moghadas, Seyed M. and Viches, Thomas and Galvani, Alison P. , title =. 2021 , month =

    2021

  6. [6]

    2025 , month =

    Gonzalez, Dulce and Karpman, Michael and Gupta, Poonam and Waxman, Elaine , title =. 2025 , month =

    2025

  7. [7]

    Journal of community health , volume=

    Unreliable and difficult-to-access food for those in need: A qualitative and quantitative study of urban food pantries , author=. Journal of community health , volume=. 2019 , publisher=

    2019

  8. [8]

    arXiv preprint arXiv:2508.21753 , year=

    Sequential Fair Allocation With Replenishments: A Little Envy Goes An Exponentially Long Way , author=. arXiv preprint arXiv:2508.21753 , year=

    work page arXiv

  9. [9]

    , title=

    Varian, Hal R. , title=. Journal of Economic Theory , year=1974, volume=. doi:, abstract=

    1974

  10. [10]

    Journal of Human Sciences and Extension , volume=

    Using a Mobile Food Pantry to Address the Food Insecurity Needs of College Students , author=. Journal of Human Sciences and Extension , volume=. 2023 , publisher=

    2023

  11. [11]

    The Washington Post , year =

    Desperate for medical equipment, states encounter a beleaguered national stockpile , author =. The Washington Post , year =

  12. [12]

    Transportation Science , volume=

    Dynamic allocations for multi-product distribution , author=. Transportation Science , volume=. 1995 , publisher=

    1995

  13. [13]

    Management Science , volume=

    Price discrimination with fairness constraints , author=. Management Science , volume=. 2022 , publisher=

    2022

  14. [14]

    Management Science , volume=

    Risk-pooling along a fixed delivery route using a dynamic inventory-allocation policy , author=. Management Science , volume=. 1995 , publisher=

    1995

  15. [15]

    Operations Research , volume=

    Sequential resource allocation for nonprofit operations , author=. Operations Research , volume=. 2014 , publisher=

    2014

  16. [16]

    , title =

    Banerjee, Siddhartha and Hssaine, Chamsi and Sinclair, Sean R. , title =. SIGMETRICS Perform. Eval. Rev. , month =. 2023 , issue_date =. doi:10.1145/3606376.3593558 , abstract =

    work page doi:10.1145/3606376.3593558 2023

  17. [17]

    and Ruiz, Angel , title =

    Rivera, Adrian Fernando and Smith, Neale R. and Ruiz, Angel , title =. Journal of Humanitarian Logistics and Supply Chain Management , year =. doi:10.1108/JHLSCM-09-2021-0087 , url =

    work page doi:10.1108/jhlscm-09-2021-0087 2021

  18. [18]

    Food bank operations: review of operations research methods and challenges during COVID-19 , journal =

    Esmaeilidouki, Aida and Rambe, Mohana and Ardestani. Food bank operations: review of operations research methods and challenges during COVID-19 , journal =. 2023 , volume =. doi:10.1186/s12889-023-16269-4 , url =

    work page doi:10.1186/s12889-023-16269-4 2023

  19. [19]

    , title =

    Nagurney, Anna and Nagurney, Ladimer S. , title =. Dynamics of Disasters: Key Concepts, Models, Algorithms, and Insights , editor =. 2016 , pages =

    2016

  20. [20]

    2019 , url =

    Olanrewaju, Oluwasegun and Hu, Shaolong and Dong, Zhijie Sasha , title =. 2019 , url =

    2019

  21. [21]

    Annals of Operations Research , volume=

    Reliable design of humanitarian supply chain under correlated disruptions: a two-stage distributionally robust approach , author=. Annals of Operations Research , volume=. 2025 , publisher=

    2025

  22. [22]

    and Diabat, Ali , title =

    Alkaabneh, Faisal and Shehadeh, Karmel S. and Diabat, Ali , title =. 2022 , url =

    2022

  23. [23]

    , title =

    Banerjee, Siddhartha and Hssaine, Chamsi and Sinclair, Sean R. , title =. Abstract Proceedings of the 2023 ACM SIGMETRICS International Conference on Measurement and Modeling of Computer Systems , pages =. 2023 , isbn =. doi:10.1145/3578338.3593558 , abstract =

    work page doi:10.1145/3578338.3593558 2023

  24. [24]

    Production and Operations Management , volume=

    On fairness and efficiency in nonprofit operations: Dynamic resource allocations , author=. Production and Operations Management , volume=. 2023 , publisher=

    2023

  25. [25]

    Proceedings of the 22nd ACM Conference on Economics and Computation , pages=

    Fair dynamic rationing , author=. Proceedings of the 22nd ACM Conference on Economics and Computation , pages=

  26. [26]

    ACM SIGMETRICS Performance Evaluation Review , volume=

    Sequential fair allocation: Achieving the optimal envy-efficiency tradeoff curve , author=. ACM SIGMETRICS Performance Evaluation Review , volume=. 2022 , publisher=

    2022

  27. [27]

    OR Spectrum , volume=

    Inventory routing for the last mile delivery of humanitarian relief supplies , author=. OR Spectrum , volume=. 2020 , publisher=

    2020

  28. [28]

    Operations Management Research , volume=

    A multiagent approach to solving the dynamic postdisaster relief distribution problem , author=. Operations Management Research , volume=. 2021 , publisher=

    2021

  29. [29]

    Davis and Irem Sengul and Julie S

    Lauren B. Davis and Irem Sengul and Julie S. Ivy and Luther G. Brock and Lastella Miles , doi =. Scheduling food bank collections and deliveries to ensure food safety and improve access , url =. Socio-Economic Planning Sciences , keywords =. 2014 , bdsk-url-1 =

    2014

  30. [30]

    European Journal of Operational Research , volume=

    Optimal two-phase vaccine allocation to geographically different regions under uncertainty , author=. European Journal of Operational Research , volume=. 2014 , publisher=

    2014

  31. [31]

    Nonprofit Operations and Supply Chain Management: Theory and Practice , pages=

    Food bank supply chain and operations in a data-driven world: objectives, challenges, and opportunities , author=. Nonprofit Operations and Supply Chain Management: Theory and Practice , pages=. 2025 , publisher=

    2025

  32. [32]

    Advances in managing humanitarian operations , pages=

    Achieving equity, effectiveness, and efficiency in food bank operations: Strategies for feeding America with implications for global hunger relief , author=. Advances in managing humanitarian operations , pages=. 2016 , publisher=

    2016

  33. [33]

    Production and Operations Management , volume=

    Elicitation of preference among multiple criteria in food distribution by food banks , author=. Production and Operations Management , volume=. 2021 , publisher=

    2021

  34. [34]

    OR Spectrum , volume=

    Modeling the role of efficiency for the equitable and effective distribution of donated food , author=. OR Spectrum , volume=. 2022 , publisher=

    2022

  35. [35]

    2022 , howpublished =

    Inflation taking toll on food banks as more Chicagoland families face hunger , author =. 2022 , howpublished =

    2022

  36. [36]

    2020 , url =

    National Council of Nonprofits , title =. 2020 , url =

    2020

  37. [37]

    2024 , url =

    Feeding America , title =. 2024 , url =

    2024

  38. [38]

    2023 , howpublished =

    Millions of American families struggle to get food on the table, report finds , author =. 2023 , howpublished =

    2023

  39. [39]

    Operations Research , volume=

    The price of fairness , author=. Operations Research , volume=. 2011 , publisher=

    2011

  40. [40]

    Operations Research , volume=

    Fairness, efficiency, and flexibility in organ allocation for kidney transplantation , author=. Operations Research , volume=. 2013 , publisher=

    2013

  41. [41]

    Proceedings of the Twenty-Eighth International Joint Conference on Artificial Intelligence,

    Fair Online Allocation of Perishable Goods and its Application to Electric Vehicle Charging , author =. Proceedings of the Twenty-Eighth International Joint Conference on Artificial Intelligence,. 2019 , month =. doi:10.24963/ijcai.2019/773 , url =

    work page doi:10.24963/ijcai.2019/773 2019

  42. [42]

    IEEE Transactions on Smart Grid , volume=

    Adaptive real power capping method for fair overvoltage regulation of distribution networks with high penetration of PV systems , author=. IEEE Transactions on Smart Grid , volume=. 2014 , publisher=

    2014

  43. [43]

    8th USENIX symposium on networked systems design and implementation (NSDI 11) , year=

    Dominant resource fairness: Fair allocation of multiple resource types , author=. 8th USENIX symposium on networked systems design and implementation (NSDI 11) , year=

  44. [44]

    Transportation science , volume=

    Thirty years of inventory routing , author=. Transportation science , volume=. 2014 , publisher=

    2014

  45. [45]

    Interfaces , volume=

    Improving the distribution of industrial gases with an on-line computerized routing and scheduling optimizer , author=. Interfaces , volume=. 1983 , publisher=

    1983

  46. [46]

    Networks , volume=

    A generalized assignment heuristic for vehicle routing , author=. Networks , volume=. 1981 , publisher=

    1981

  47. [47]

    Transportation Research Part B: Methodological , volume=

    Analyzing trade-offs between transportation, inventory and production costs on freight networks , author=. Transportation Research Part B: Methodological , volume=. 1985 , publisher=

    1985

  48. [48]

    Management science , volume=

    One warehouse multiple retailer systems with vehicle routing costs , author=. Management science , volume=. 1990 , publisher=

    1990

  49. [49]

    Annals of Operations Research , volume=

    Vehicle routing problems over time: a survey , author=. Annals of Operations Research , volume=. 2022 , publisher=

    2022

  50. [50]

    Transportation Science , volume=

    Large neighborhoods with implicit customer selection for vehicle routing problems with profits , author=. Transportation Science , volume=. 2016 , publisher=

    2016

  51. [51]

    International transactions in operational research , volume=

    Vehicle routing problems with split deliveries , author=. International transactions in operational research , volume=. 2012 , publisher=

    2012

  52. [52]

    European Journal of Operational Research , volume=

    Multi-compartment vehicle routing problems: State-of-the-art, modeling framework and future directions , author=. European Journal of Operational Research , volume=. 2021 , publisher=

    2021

  53. [53]

    Transportation Science , volume=

    Deliveries in an inventory/routing problem using stochastic dynamic programming , author=. Transportation Science , volume=. 2001 , publisher=

    2001

  54. [54]

    Transportation science , volume=

    The stochastic inventory routing problem with direct deliveries , author=. Transportation science , volume=. 2002 , publisher=

    2002

  55. [55]

    Operations Research , volume=

    Indexability and index heuristics for a simple class of inventory routing problems , author=. Operations Research , volume=. 2009 , publisher=

    2009

  56. [56]

    Transportation Research Part C: Emerging Technologies , volume=

    A stochastic inventory routing problem with stock-out , author=. Transportation Research Part C: Emerging Technologies , volume=. 2013 , publisher=

    2013

  57. [57]

    Computers & Operations Research , volume=

    Stochastic inventory routing with dynamic demands and intra-day depletion , author=. Computers & Operations Research , volume=. 2024 , publisher=

    2024

  58. [58]

    Proceedings of the 2020 conference on fairness, accountability, and transparency , pages=

    Fairness and utilization in allocating resources with uncertain demand , author=. Proceedings of the 2020 conference on fairness, accountability, and transparency , pages=

    2020

  59. [59]

    Proceedings of the 21st ACM Conference on Economics and Computation , pages=

    EFX exists for three agents , author=. Proceedings of the 21st ACM Conference on Economics and Computation , pages=

  60. [60]

    Available at SSRN 3646539 , year=

    Effective, fair and equitable pandemic rationing , author=. Available at SSRN 3646539 , year=

  61. [61]

    Operations Research , volume=

    The humanitarian pickup and distribution problem , author=. Operations Research , volume=. 2019 , publisher=

    2019

  62. [62]

    Transportation Research Part C: Emerging Technologies , volume=

    Routing and resource allocation in non-profit settings with equity and efficiency measures under demand uncertainty , author=. Transportation Research Part C: Emerging Technologies , volume=. 2023 , publisher=

    2023

  63. [63]

    IIE Transactions , volume=

    Multi-vehicle sequential resource allocation for a nonprofit distribution system , author=. IIE Transactions , volume=. 2014 , publisher=

    2014

  64. [64]

    2024 , school=

    Strategies for Maintaining the Financial Sustainability of Nonprofit Organizations During Economic Downturns , author=. 2024 , school=

    2024

  65. [65]

    VOLUNTAS: International Journal of Voluntary and Nonprofit Organizations , volume=

    The slaughtered and the survivors: Collaboration between social economy organizations as a key to success in times of financial crisis , author=. VOLUNTAS: International Journal of Voluntary and Nonprofit Organizations , volume=. 2017 , publisher=

    2017

  66. [66]

    The Social science journal , volume=

    Sustainability of social economy organizations (SEOs): An analysis of the conditions for surviving and thriving , author=. The Social science journal , volume=. 2024 , publisher=

    2024

  67. [67]

    2023 , url =

    Global Citizen , title =. 2023 , url =

    2023

  68. [68]

    2024 Annual Report , year =

    2024

  69. [69]

    Operations Research , volume=

    The price of fairness of scheduling a scarce resource , author=. Operations Research , volume=. 2025 , publisher=

    2025

  70. [70]

    Informatics , volume=

    Issues in multiagent resource allocation , author=. Informatics , volume=. 2005 , publisher=

    2005

  71. [71]

    Theory of Computing Systems , volume=

    The efficiency of fair division , author=. Theory of Computing Systems , volume=. 2012 , publisher=

    2012

  72. [72]

    Operations Research , volume=

    A combined vehicle routing and inventory allocation problem , author=. Operations Research , volume=. 1984 , publisher=

    1984

  73. [73]

    Operations Research , volume=

    Distribution strategies that minimize transportation and inventory costs , author=. Operations Research , volume=. 1985 , publisher=

    1985

  74. [74]

    International Conference on Web and Internet Economics , pages=

    On best-of-both-worlds fair-share allocations , author=. International Conference on Web and Internet Economics , pages=. 2022 , organization=

    2022

  75. [75]

    Proceedings of the twenty-seventh annual ACM-SIAM symposium on Discrete algorithms , pages=

    Online contention resolution schemes , author=. Proceedings of the twenty-seventh annual ACM-SIAM symposium on Discrete algorithms , pages=. 2016 , organization=

    2016

  76. [76]

    Operations Research , volume=

    On (random-order) online contention resolution schemes for the matching polytope of (bipartite) graphs , author=. Operations Research , volume=. 2025 , publisher=

    2025

  77. [77]

    arXiv preprint arXiv:2403.05378 , year=

    Online contention resolution schemes for network revenue management and combinatorial auctions , author=. arXiv preprint arXiv:2403.05378 , year=

    work page arXiv

  78. [78]

    Proceedings of the 26th ACM Conference on Economics and Computation , pages=

    Forward-backward Contention Resolution Schemes for Fair Rationing , author=. Proceedings of the 26th ACM Conference on Economics and Computation , pages=

  79. [79]

    Management Science , year=

    Markovian Search with Ex Ante Constraints: Theory and Applications to Socially Aware Algorithmic Hiring , author=. Management Science , year=

  80. [80]

    Proceedings of the Thirty-Third International Joint Conference on Artificial Intelligence , pages=

    Online combinatorial optimization with group fairness constraints , author=. Proceedings of the Thirty-Third International Joint Conference on Artificial Intelligence , pages=

Showing first 80 references.