pith. sign in

arxiv: 2409.08347 · v5 · submitted 2024-09-12 · 💰 econ.EM · cs.GT· math.OC

Sensitivity analysis of the perturbed utility stochastic traffic equilibrium

Mogens Fosgerau , Nikolaj Nielsen , Mads Paulsen , Thomas Kj{\ae}r Rasmussen , Rui Yao This is my paper

Pith reviewed 2026-05-23 20:44 UTC · model grok-4.3

classification 💰 econ.EM cs.GTmath.OC
keywords sensitivity analysisperturbed utility route choicestochastic traffic equilibriumJacobianlink flowslink coststransportation planningequilibrium sensitivity
0
0 comments X

The pith

Analytical expressions are derived for the Jacobian of equilibrium link flows with respect to link costs in the perturbed utility stochastic traffic model.

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

This paper develops a sensitivity analysis framework for the perturbed utility route choice model and its associated stochastic traffic equilibrium. It derives analytical expressions for the Jacobian of individual optimal flows and equilibrium link flows with respect to link cost parameters. The expressions hold under general assumptions and allow computation of marginal flow changes from marginal cost changes across a network. Implementation exploits sparsity in the model, and the results support applications such as network design evaluation and uncertainty quantification in transportation planning.

Core claim

The central claim is that analytical sensitivity expressions can be derived for the Jacobian of the individual optimal PURC flow and equilibrium link flows with respect to link cost parameters under general assumptions. This permits determining the marginal change in link flows following a marginal change in link costs across the network, with implementation that exploits the sparsity generated by the PURC model.

What carries the argument

Analytical sensitivity expressions obtained through differentiation of the optimality conditions of the PURC model and the stochastic equilibrium with respect to link costs.

If this is right

  • Marginal changes in equilibrium link flows can be computed directly from the Jacobian without re-solving the equilibrium problem.
  • Sparsity from the PURC model enables efficient implementation of the sensitivities even for large-scale networks.
  • Critical design parameters can be identified through examination of the sensitivity values.
  • Uncertainty in performance predictions can be quantified using the derived sensitivities.
  • The results apply directly to network design, pricing strategies, and policy analysis.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The approach may extend to other stochastic equilibrium models provided they permit similar closed-form differentiation.
  • The sensitivities could support gradient-based optimization routines for network design problems.
  • Integration with real-time data streams might enable dynamic adjustment of flow predictions following observed cost changes.

Load-bearing premise

The PURC model and stochastic equilibrium admit closed-form differentiation of the optimality conditions with respect to link costs.

What would settle it

A numerical verification on a small test network comparing the analytical Jacobian values to finite-difference approximations of the equilibrium link flows after perturbing a link cost; systematic mismatch would falsify the claimed derivations.

read the original abstract

This paper develops a sensitivity analysis framework for the perturbed utility route choice (PURC) model and the accompanying stochastic traffic equilibrium model. We derive analytical sensitivity expressions for the Jacobian of the individual optimal PURC flow and equilibrium link flows with respect to link cost parameters under general assumptions. This allows us to determine the marginal change in link flows following a marginal change in link costs across the network. We show how to implement these results while exploiting the sparsity generated by the PURC model. Numerical examples illustrate the use of our method for estimating equilibrium link flows after link cost shifts, identifying critical design parameters, and quantifying uncertainty in performance predictions. Finally, we demonstrate the method in a large-scale example. The findings have implications for network design, pricing strategies, and policy analysis in transportation planning and economics, providing a bridge between theoretical models and real-world applications.

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

1 major / 0 minor

Summary. The paper develops a sensitivity analysis framework for the perturbed utility route choice (PURC) model and the associated stochastic traffic equilibrium. It derives analytical expressions for the Jacobian of the individual optimal PURC flows and equilibrium link flows with respect to link cost parameters under general assumptions, shows how to implement these while exploiting sparsity, and illustrates the approach with numerical examples on small and large-scale networks for applications including post-shift flow estimation, critical parameter identification, and uncertainty quantification.

Significance. If the derivations are valid, the framework supplies a computationally efficient way to obtain marginal effects of cost perturbations on equilibrium flows without re-solving the full equilibrium problem each time. This is useful for network design, pricing, and policy analysis in transportation. The emphasis on sparsity exploitation and large-scale demonstration are practical strengths that could facilitate adoption in applied work.

major comments (1)
  1. [Abstract / derivation of sensitivities] Abstract and the derivation section: the central claim is that closed-form Jacobians exist 'under general assumptions.' Application of the implicit function theorem to the first-order conditions or fixed-point map of the stochastic equilibrium requires (at minimum) continuous differentiability of the choice probabilities and nonsingularity of the relevant Jacobian, which in turn needs strict convexity or strong monotonicity conditions on the perturbation distribution and link costs. These regularity conditions are not stated, so the generality claim is not yet load-bearing for all networks (e.g., those with flat cost regions).

Simulated Author's Rebuttal

1 responses · 0 unresolved

We thank the referee for the constructive feedback and positive evaluation of the paper's practical contributions. We address the single major comment below.

read point-by-point responses

  1. Referee: [Abstract / derivation of sensitivities] Abstract and the derivation section: the central claim is that closed-form Jacobians exist 'under general assumptions.' Application of the implicit function theorem to the first-order conditions or fixed-point map of the stochastic equilibrium requires (at minimum) continuous differentiability of the choice probabilities and nonsingularity of the relevant Jacobian, which in turn needs strict convexity or strong monotonicity conditions on the perturbation distribution and link costs. These regularity conditions are not stated, so the generality claim is not yet load-bearing for all networks (e.g., those with flat cost regions).

    Authors: We agree that the regularity conditions supporting the implicit function theorem application should be stated explicitly rather than left implicit under the phrase 'general assumptions.' In the revision we will insert a short paragraph (or subsection) immediately preceding the main derivations that lists the minimal conditions: (i) continuous differentiability of the PURC choice probabilities, which follows from standard assumptions on the perturbation distribution (positive density on R and strict convexity of the perturbation function), and (ii) nonsingularity of the relevant Jacobian of the equilibrium fixed-point map, ensured by strict monotonicity of link costs together with the strict convexity already imposed on the perturbation. These conditions rule out degenerate cases such as flat cost regions. We will also revise the abstract to read 'under the regularity conditions stated in Section X' so that the generality claim is properly qualified. revision: yes

Circularity Check

0 steps flagged

No circularity; direct differentiation of equilibrium conditions under stated assumptions

full rationale

The paper's core claim is derivation of analytical Jacobians for PURC flows and equilibrium link flows w.r.t. link costs via differentiation of optimality conditions. No quoted equations or self-citations reduce the result to a fitted input, renamed pattern, or self-referential definition. The approach is a standard application of implicit differentiation to a fixed-point map, presented as self-contained under general assumptions without load-bearing reliance on prior author work that would create circularity. This matches the default expectation of non-circularity for such sensitivity analyses.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

Only abstract available; ledger entries are inferred from stated claims. The central derivations rest on the existence of differentiable optimality conditions under unspecified general assumptions.

axioms (1)
  • domain assumption PURC model and stochastic equilibrium admit analytical Jacobians w.r.t. link costs under general assumptions
    Invoked directly in the abstract as the basis for the sensitivity expressions

pith-pipeline@v0.9.0 · 5686 in / 1183 out tokens · 22064 ms · 2026-05-23T20:44:58.730082+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

68 extracted references · 68 canonical work pages

  1. [1]

    , " * write output.state after.block = add.period write newline

    ENTRY address author booktitle chapter edition editor howpublished institution journal key month note number organization pages publisher school series title type volume year label extra.label sort.label short.list INTEGERS output.state before.all mid.sentence after.sentence after.block FUNCTION init.state.consts #0 'before.all := #1 'mid.sentence := #2 '...

    work page

  2. [2]

    write newline

    " write newline "" before.all 'output.state := FUNCTION n.dashify 't := "" t empty not t #1 #1 substring "-" = t #1 #2 substring "--" = not "--" * t #2 global.max substring 't := t #1 #1 substring "-" = "-" * t #2 global.max substring 't := while if t #1 #1 substring * t #2 global.max substring 't := if while FUNCTION word.in bbl.in capitalize ":" * " " *...

    work page

  3. [3]

    D., 1939

    Aleksandrov, A. D., 1939. Almost everywhere existence of the second differential of a convex function and some properties of convex surfaces connected with it. Uchenye Zapiski Leningrad Gos. Univ., Math. Ser 6, 3--35

    work page 1939

  4. [4]

    Identification with additively separable heterogeneity

    Allen, R., Rehbeck, J., 2019. Identification with additively separable heterogeneity. Econometrica 87, 1021--1054

    work page 2019

  5. [5]

    Z., Koltun, V., 2019

    Bai, S., Kolter, J. Z., Koltun, V., 2019. Deep equilibrium models. Advances in neural information processing systems 32

    work page 2019

  6. [6]

    Markovian traffic equilibrium

    Baillon, J.-B., Cominetti, R., 2008. Markovian traffic equilibrium. Mathematical Programming 111, 33--56

    work page 2008

  7. [7]

    Stochastic user equilibrium formulation for generalized nested logit model

    Bekhor, S., Prashker, J., 2001. Stochastic user equilibrium formulation for generalized nested logit model. Transportation Research Record 1752, 84--90

    work page 2001

  8. [8]

    N., 1999

    Bekhor, S., Prashker, J. N., 1999. Formulations of extended logit stochastic user equilibrium assignments. In: 14th International Symposium on Transportation and Traffic TheoryTransportation Research Institute\/

    work page 1999

  9. [9]

    G., Cassir, C., Iida, Y., Lam, W

    Bell, M. G., Cassir, C., Iida, Y., Lam, W. H., 1999. A sensitivity based approach to network reliability assessment. In: 14th International Symposium on Transportation and Traffic TheoryTransportation Research Institute\/

    work page 1999

  10. [10]

    Discrete choice methods and their applications to short term travel decisions

    Ben-Akiva, M., Bierlaire, M., 1999. Discrete choice methods and their applications to short term travel decisions. In: Handbook of transportation science\/ , Springer, pp. 5--33

    work page 1999

  11. [11]

    E., Lerman, S

    Ben-Akiva, M. E., Lerman, S. R., 1985. Discrete choice analysis: theory and application to travel demand, vol. 9. MIT press

    work page 1985

  12. [12]

    H., Bekhor, S., Prato, C

    Bovy, P. H., Bekhor, S., Prato, C. G., 2008. The factor of revisited path size: Alternative derivation. Transportation Research Record 2076, 132--140

    work page 2008

  13. [13]

    D., 2012

    Boyles, S. D., 2012. Bush-based sensitivity analysis for approximating subnetwork diversion. Transportation Research Part B: Methodological 46, 139--155

    work page 2012

  14. [14]

    E., Friesz, T

    Cho, H.-J., Smith, T. E., Friesz, T. L., 2000. A reduction method for local sensitivity analyses of network equilibrium arc flows. Transportation Research Part B: Methodological 34, 31--51

    work page 2000

  15. [15]

    A paired combinatorial logit model for travel demand analysis

    Chu, C., 1989. A paired combinatorial logit model for travel demand analysis. In: Proceedings of the fifth world conference on transportation research\/ , vol. 4, pp. 295--309

    work page 1989

  16. [16]

    D., Cho, H.-J., Friesz, T

    Chung, B. D., Cho, H.-J., Friesz, T. L., Huang, H., Yao, T., 2014. Sensitivity analysis of user equilibrium flows revisited. Networks and Spatial Economics 14, 183--207

    work page 2014

  17. [17]

    D., Watling, D

    Clark, S. D., Watling, D. P., 2002. Sensitivity analysis of the probit-based stochastic user equilibrium assignment model. Transportation Research Part B: Methodological 36, 617--635

    work page 2002

  18. [18]

    D., Watling, D

    Clark, S. D., Watling, D. P., 2006. Applications of sensitivity analysis for probit stochastic network equilibrium. European journal of operational research 175, 894--911

    work page 2006

  19. [19]

    H., 1990

    Clarke, F. H., 1990. Optimization and nonsmooth analysis. SIAM

    work page 1990

  20. [20]

    B., 1971

    Dial, R. B., 1971. A probabilistic multipath traffic assignment model which obviates path enumeration. Transportation research 5, 83--111

    work page 1971

  21. [21]

    C., Watling, D

    Duncan, L. C., Watling, D. P., Connors, R. D., Rasmussen, T. K., Nielsen, O. A., 2020. Path size logit route choice models: Issues with current models, a new internally consistent approach, and parameter estimation on a large-scale network with gps data. Transportation Research Part B: Methodological 135, 1--40

    work page 2020

  22. [22]

    A link based network route choice model with unrestricted choice set

    Fosgerau, M., Frejinger, E., Karlstrom, A., 2013 a . A link based network route choice model with unrestricted choice set. Transportation Research Part B: Methodological 56, 70--80

    work page 2013

  23. [23]

    Choice probability generating functions

    Fosgerau, M., McFadden, D., Bierlaire, M., 2013 b . Choice probability generating functions. Journal of Choice Modelling 8

    work page 2013

  24. [24]

    L., 2012

    Fosgerau, M., McFadden, D. L., 2012. A theory of the perturbed consumer with general budgets. NBER Working Paper 17953

    work page 2012

  25. [25]

    R.-V., 2021

    Fosgerau, M., Melo, E., Shum, M., Sørensen, J. R.-V., 2021. Some Remarks on CCP -based Estimators of Dynamic Models . Economics Letters 204

    work page 2021

  26. [26]

    The Inverse Product Differentiation Logit Model

    Fosgerau, M., Monardo, J., de Palma, A., 2024. The Inverse Product Differentiation Logit Model . American Economic Journal: Microeconomics 16, 329--70

    work page 2024

  27. [27]

    K., 2022

    Fosgerau, M., Paulsen, M., Rasmussen, T. K., 2022. A perturbed utility route choice model. Transportation Research Part C 136, arXiv: 2103.13784v1 ISBN: 2103.13784v1 Publisher: arXiv

    work page arXiv 2022

  28. [28]

    Bikeability and the induced demand for cycling

    Fosgerau, M., Łukawska, M., Paulsen, M., Rasmussen, T., 2023. Bikeability and the induced demand for cycling. Proceedings of the National Academy of Sciences

    work page 2023

  29. [29]

    Bilevel programming for hyperparameter optimization and meta-learning

    Franceschi, L., Frasconi, P., Salzo, S., Grazzi, R., Pontil, M., 2018. Bilevel programming for hyperparameter optimization and meta-learning. In: International conference on machine learning\/ , PMLR, pp. 1568--1577

    work page 2018

  30. [30]

    Route choice using a paired combinatorial logit model

    Gliebe, J., Koppelman, F., Ziliaskopoulos, A., 1999. Route choice using a paired combinatorial logit model. In: 78th TRB Meeting, Washington DC\/

    work page 1999

  31. [31]

    H., 2002

    Hofbauer, J., Sandholm, W. H., 2002. On the global convergence of stochastic fictitious play. Econometrica 70, 2265--2294

    work page 2002

  32. [32]

    D., 2016

    Jafari, E., Boyles, S. D., 2016. Improved bush-based methods for network contraction. Transportation Research Part B: Methodological 83, 298--313

    work page 2016

  33. [33]

    Sensitivity analysis of separable traffic equilibrium equilibria with application to bilevel optimization in network design

    Josefsson, M., Patriksson, M., 2007. Sensitivity analysis of separable traffic equilibrium equilibria with application to bilevel optimization in network design. Transportation Research Part B: Methodological 41, 4--31

    work page 2007

  34. [34]

    Unconstrained weibit stochastic user equilibrium model with extensions

    Kitthamkesorn, S., Chen, A., 2014. Unconstrained weibit stochastic user equilibrium model with extensions. Transportation Research Part B: Methodological 59, 1--21

    work page 2014

  35. [35]

    G., Parks, H

    Krantz, S. G., Parks, H. R., 2003. The Implicit Function Theorem . Birkhäuser, Boston, MA

    work page 2003

  36. [36]

    Solution differentiability for variational inequalities

    Kyparisis, J., 1990. Solution differentiability for variational inequalities. Mathematical Programming 48, 285--301

    work page 1990

  37. [37]

    Inducing equilibria via incentives: Simultaneous design-and-play ensures global convergence

    Liu, B., Li, J., Yang, Z., Wai, H.-T., Hong, M., Nie, Y., Wang, Z., 2022. Inducing equilibria via incentives: Simultaneous design-and-play ensures global convergence. Advances in Neural Information Processing Systems 35, 29001--29013

    work page 2022

  38. [38]

    Optimizing millions of hyperparameters by implicit differentiation

    Lorraine, J., Vicol, P., Duvenaud, D., 2020. Optimizing millions of hyperparameters by implicit differentiation. In: International conference on artificial intelligence and statistics\/ , PMLR, pp. 1540--1552

    work page 2020

  39. [39]

    A probit-based stochastic user equilibrium assignment model

    Maher, M., Hughes, P., 1997. A probit-based stochastic user equilibrium assignment model. Transportation Research Part B: Methodological 31, 341--355

    work page 1997

  40. [40]

    A method of integrating correlation structures for a generalized recursive route choice model

    Mai, T., 2016. A method of integrating correlation structures for a generalized recursive route choice model. Transportation Research Part B: Methodological 93, 146--161

    work page 2016

  41. [41]

    A nested recursive logit model for route choice analysis

    Mai, T., Fosgerau, M., Frejinger, E., 2015. A nested recursive logit model for route choice analysis. Transportation Research Part B: Methodological 75, 100--112

    work page 2015

  42. [42]

    On the covariance structure of the cross-nested logit model

    Marzano, V., Papola, A., 2008. On the covariance structure of the cross-nested logit model. Transportation Research Part B: Methodological 42, 83--98

    work page 2008

  43. [43]

    Econometric Models of Probabilistic Choice

    McFadden, D., 1981. Econometric Models of Probabilistic Choice . In: Manski, C., McFadden, D. (eds.), Structural Analysis of Discrete Data with Econometric Applications \/ , MIT Press, Cambridge, MA, USA, pp. 198--272, iSSN: 00219398

    work page 1981

  44. [44]

    Envelope Theorems for Arbitrary Choice Sets

    Milgrom, P., Segal, I., 2002. Envelope Theorems for Arbitrary Choice Sets . Econometrica 70, 583--601, iSBN: 00129682

    work page 2002

  45. [45]

    Markovian traffic equilibrium assignment based on network generalized extreme value model

    Oyama, Y., Hara, Y., Akamatsu, T., 2022. Markovian traffic equilibrium assignment based on network generalized extreme value model. Transportation Research Part B: Methodological 155, 135--159

    work page 2022

  46. [46]

    Prism-based path set restriction for solving markovian traffic assignment problem

    Oyama, Y., Hato, E., 2019. Prism-based path set restriction for solving markovian traffic assignment problem. Transportation Research Part B: Methodological 122, 528--546

    work page 2019

  47. [47]

    Sensitivity Analysis of Traffic Equilibria

    Patriksson, M., 2004. Sensitivity Analysis of Traffic Equilibria . Transportation Science 38, 258--281, publisher: INFORMS

    work page 2004

  48. [48]

    T., 2003

    Patriksson, M., Rockafellar, R. T., 2003. Sensitivity analysis of aggregated variational inequality problems, with application to traffic equilibria. Transportation Science 37, 56--68

    work page 2003

  49. [49]

    L., 1989

    Qiu, Y., Magnanti, T. L., 1989. Sensitivity analysis for variational inequalities defined on polyhedral sets. Mathematics of Operations Research 14, 410--432

    work page 1989

  50. [50]

    S., 2001

    Ramming, M. S., 2001. Network knowledge and route choice. Unpublished Ph. D. Thesis, Massachusetts Institute of Technology

    work page 2001

  51. [51]

    Estimating the Number of s-t Paths in a Graph

    Roberts, B., Kroese, D., 2007. Estimating the Number of s-t Paths in a Graph . Journal of Graph Algorithms and Applications 11, 195--214

    work page 2007

  52. [52]

    M., 1980

    Robinson, S. M., 1980. Strongly Regular Generalized Equations . Mathematics of Operations Research 5, 43--62

    work page 1980

  53. [53]

    M., 1985

    Robinson, S. M., 1985. Implicit b-differentiability in generalized equations(technical summary report)

    work page 1985

  54. [54]

    M., 2006

    Robinson, S. M., 2006. Strong regularity and the sensitivity analysis of traffic equilibria: A comment. Transportation Science 40, 540--542

    work page 2006

  55. [55]

    T., 1970

    Rockafellar, R. T., 1970. Convex Analysis . Princeton University Press, Princeton, N.J

    work page 1970

  56. [56]

    Urban transportation networks, vol

    Sheffi, Y., 1985. Urban transportation networks, vol. 6. Prentice-Hall, Englewood Cliffs, NJ

    work page 1985

  57. [57]

    R.-V., Fosgerau, M., 2022

    Sørensen, J. R.-V., Fosgerau, M., 2022. How McFadden met Rockafellar and learned to do more with less. Journal of Mathematical Economics 100

    work page 2022

  58. [58]

    L., Friesz, T

    Tobin, R. L., Friesz, T. L., 1988. Sensitivity analysis for equilibrium network flow. Transportation Science 22, 242--250

    work page 1988

  59. [59]

    Application of cross-nested logit model to mode choice in tel aviv, israel, metropolitan area

    Vovsha, P., 1997. Application of cross-nested logit model to mode choice in tel aviv, israel, metropolitan area. Transportation Research Record 1607, 6--15

    work page 1997

  60. [60]

    The Link - Nested Logit model of route choice: overcoming the route overlapping problem

    Vovsha, P., Bekhor, S., 1998. The Link - Nested Logit model of route choice: overcoming the route overlapping problem. Transportation Research Record 1645, 133--142

    work page 1998

  61. [61]

    S., 2001

    Wen, C.-H., Koppelman, F. S., 2001. The generalized nested logit model. Transportation Research Part B: Methodological 35, 627--641

    work page 2001

  62. [62]

    Sensitivity analysis of the combined travel demand model with applications

    Yang, C., Chen, A., 2009. Sensitivity analysis of the combined travel demand model with applications. European Journal of Operational Research 198, 909--921

    work page 2009

  63. [63]

    Sensitivity-based uncertainty analysis of a combined travel demand model

    Yang, C., Chen, A., Xu, X., Wong, S., 2013. Sensitivity-based uncertainty analysis of a combined travel demand model. Transportation Research Part B: Methodological 57, 225--244

    work page 2013

  64. [64]

    G., 1997

    Yang, H., Bell, M. G., 1997. Traffic restraint, road pricing and network equilibrium. Transportation Research Part B: Methodological 31, 303--314

    work page 1997

  65. [65]

    G., 2007

    Yang, H., Bell, M. G., 2007. Sensitivity analysis of network traffic equilibrium revisited: the corrected approach. In: 4th IMA International Conference on Mathematics in TransportInstitute of Mathematics and its Applications\/

    work page 2007

  66. [66]

    Perturbed utility stochastic traffic assignment

    Yao, R., Fosgerau, M., Paulsen, M., Rasmussen, T., 2024. Perturbed utility stochastic traffic assignment. Transportation Science 58

    work page 2024

  67. [67]

    Q., Miyagi, T., 2001

    Ying, J. Q., Miyagi, T., 2001. Sensitivity analysis for stochastic user equilibrium network flows—a dual approach. Transportation Science 35, 124--133

    work page 2001

  68. [68]

    Q., Yang, H., 2005

    Ying, J. Q., Yang, H., 2005. Sensitivity analysis of stochastic user equilibrium flows in a bi-modal network with application to optimal pricing. Transportation research Part B: methodological 39, 769--795

    work page 2005