pith. machine review for the scientific record. sign in

arxiv: 2604.11511 · v1 · submitted 2026-04-13 · 💻 cs.GT · cs.LG

Recognition: unknown

The Price of Ignorance: Information-Free Quotation for Data Retention in Machine Unlearning

Bin Han , Di Feng , Zexin Fang , Jie Wang , Hans D. Schotten
Authors on Pith no claims yet

Pith reviewed 2026-05-10 16:24 UTC · model grok-4.3

classification 💻 cs.GT cs.LG
keywords machine unlearningdata retention pricinginformation-free mechanismprice of ignorancesubgame-perfect Nash equilibriumGDPR compliancewelfare analysis
0
0 comments X

The pith

An ascending price quotation mechanism for data retention in machine unlearning achieves at least 99 percent of the welfare of fully informed personalized pricing without knowing any user preferences.

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

Mobile network operators must balance machine unlearning to honor data deletion rights under regulations like GDPR against the accuracy costs of removing data from models. Existing retention pricing schemes require the operator to know each user's private preferences for privacy and accuracy, which itself violates the spirit of those regulations. The paper introduces a broadcast-based ascending quotation where the server raises prices over time and users simply decide whether to sell their data at the current level. This information-free protocol is proven to induce a unique equilibrium in which users sell once, and large-scale simulations show it captures nearly all the welfare of schemes that know everything.

Core claim

The paper defines the Price of Ignorance as the welfare gap between optimal personalized pricing, which requires full knowledge of every user's parameters, and an information-free ascending quotation mechanism in which the server broadcasts rising prices and users self-select their supply without revealing preferences. Under complete information the mechanism admits a unique subgame-perfect Nash equilibrium of single-period selling. The authors prove a three-regime efficiency ordering and demonstrate through 5000 Monte Carlo runs across seven mechanisms that the information-free approach reaches at least 99 percent of benchmark welfare, while remaining robust to noise and preserving fairness

What carries the argument

The ascending quotation mechanism, in which the server broadcasts progressively higher prices and users decide their data supply based on self-selected thresholds without revealing private information.

If this is right

  • The protocol admits a unique subgame-perfect Nash equilibrium characterized by single-period selling.
  • A three-regime efficiency ordering holds between the information-free quotation and its information-intensive benchmarks.
  • The mechanism delivers noise-robust performance guarantees.
  • Fairness outcomes remain comparable to those of fully informed pricing schemes.

Where Pith is reading between the lines

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

  • The same broadcast approach could be adapted to data markets in federated learning or edge computing where preference revelation is also restricted.
  • Regulators could evaluate whether mandating ascending quotation protocols reduces compliance burdens while still protecting deletion rights.
  • Repeated-interaction versions of the mechanism might further shrink any residual welfare gap by allowing users to update decisions over time.

Load-bearing premise

Users will self-select their data supply according to the broadcast prices and play the unique subgame-perfect Nash equilibrium of single-period selling, with the welfare comparison holding under the modeled preference distributions.

What would settle it

A field experiment in which real users interact with the broadcast prices and the resulting aggregate welfare is measured against a simulated personalized optimum; if the gap consistently exceeds one percent or users sell across multiple periods, the near-zero Price of Ignorance claim would be refuted.

Figures

Figures reproduced from arXiv: 2604.11511 by Bin Han, Di Feng, Hans D. Schotten, Jie Wang, Zexin Fang.

Figure 1
Figure 1. Figure 1: Different cases of the server’s cost function. [PITH_FULL_IMAGE:figures/full_fig_p004_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Privacy-only welfare comparison across mechanisms [PITH_FULL_IMAGE:figures/full_fig_p011_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Accuracy-aware welfare comparison (θi ∼ U(0, 5)). The accuracy externality reduces welfare across all mecha￾nisms. data protection regulations and information asymmetry. CIQ assumes users possess complete information about all others and play perfectly strategic equilibria—an equally unrealistic assumption. IIQ is the only implementable mechanism among the three: it requires no private parameter knowledge … view at source ↗
Figure 5
Figure 5. Figure 5: Sensitivity of social welfare to key parameters. Each panel sweeps one parameter while holding others at defaults. [PITH_FULL_IMAGE:figures/full_fig_p013_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: Left: Fairness-efficiency tradeoff across mechanisms. Right: Free-rider characterization—mean supply fraction by [PITH_FULL_IMAGE:figures/full_fig_p014_6.png] view at source ↗
Figure 7
Figure 7. Figure 7: Convergence analysis. (a) Price increments to termina [PITH_FULL_IMAGE:figures/full_fig_p014_7.png] view at source ↗
read the original abstract

When users exercise data deletion rights under the General Data Protection Regulation (GDPR) and similar regulations, mobile network operators face a tradeoff: excessive machine unlearning degrades model accuracy and incurs retraining costs, yet existing pricing mechanisms for data retention require the server to know every user's private privacy and accuracy preferences, which is infeasible under the very regulations that motivate unlearning. We ask: what is the welfare cost of operating without this private information? We design an information-free ascending quotation mechanism where the server broadcasts progressively higher prices and users self-select their data supply, requiring no knowledge of users' parameters. Under complete information, the protocol admits a unique subgame-perfect Nash equilibrium characterized by single-period selling. We formalize the Price of Ignorance -- the welfare gap between optimal personalized pricing (which knows everything) and our information-free quotation (which knows nothing) -- and prove a three-regime efficiency ordering. Numerical evaluation across seven mechanisms and 5000 Monte Carlo runs shows that this price is near zero: the information-free mechanism achieves >=99% of the welfare of its information-intensive benchmarks, while providing noise-robust guarantees and comparable fairness.

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 / 2 minor

Summary. The paper proposes an information-free ascending quotation mechanism for data retention in machine unlearning under GDPR-like regulations. The server broadcasts progressively higher prices, enabling users to self-select their data supply without the server knowing private user preferences. Under complete information, the mechanism admits a unique subgame-perfect Nash equilibrium of single-period selling. It defines the Price of Ignorance as the welfare gap to an optimal personalized pricing benchmark, proves a three-regime efficiency ordering, and reports that numerical evaluation across seven mechanisms and 5000 Monte Carlo runs shows the information-free mechanism achieves >=99% of benchmark welfare, with noise-robust guarantees and comparable fairness.

Significance. If the central claims hold, the work provides a valuable practical advance in mechanism design for privacy-preserving machine unlearning by demonstrating that near-optimal welfare can be achieved without access to private user parameters, which is essential under data deletion regulations. The formal equilibrium characterization, three-regime ordering, and extensive Monte Carlo evaluation (5000 runs) are strengths that support the feasibility of information-free designs over information-intensive alternatives.

major comments (2)
  1. [Abstract and Equilibrium Analysis] Abstract and Equilibrium section: The unique subgame-perfect Nash equilibrium is derived under complete information, with users playing single-period selling. However, the model explicitly features users with private, heterogeneous privacy and accuracy preferences (unknown to the server), making this an incomplete-information game. Bayesian Nash equilibria may differ from the complete-information SPNE in uniqueness or outcomes, which would undermine the three-regime ordering and the >=99% welfare result from the simulations that embed this assumption.
  2. [Numerical Evaluation] Numerical Evaluation section: The claim that the Price of Ignorance is near zero (>=99% welfare) rests on 5000 Monte Carlo runs across seven mechanisms. Without explicit details on preference distributions, parameter values, data-exclusion rules, or how the complete-information equilibrium is implemented under private preferences, it is unclear whether the result is robust or sensitive to post-hoc simulation choices, as the welfare gap is defined relative to an externally optimal benchmark.
minor comments (2)
  1. [Abstract] The abstract refers to 'noise-robust guarantees' without specifying the noise model (e.g., in preferences or data); this should be clarified in the main text with a reference to the relevant section.
  2. [Numerical Evaluation] Ensure the seven mechanisms compared in the numerical section are explicitly named and defined, with any connections to prior work noted for clarity.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for their careful and constructive review, which highlights key aspects of the modeling assumptions and reproducibility. We address each major comment below with clarifications and proposed revisions.

read point-by-point responses

  1. Referee: [Abstract and Equilibrium Analysis] Abstract and Equilibrium section: The unique subgame-perfect Nash equilibrium is derived under complete information, with users playing single-period selling. However, the model explicitly features users with private, heterogeneous privacy and accuracy preferences (unknown to the server), making this an incomplete-information game. Bayesian Nash equilibria may differ from the complete-information SPNE in uniqueness or outcomes, which would undermine the three-regime ordering and the >=99% welfare result from the simulations that embed this assumption.

    Authors: We appreciate the referee highlighting the distinction between complete and incomplete information. The equilibrium analysis derives the unique SPNE under complete information to obtain a clean characterization of single-period selling as the equilibrium strategy for the ascending quotation protocol; this serves as the benchmark for defining the Price of Ignorance and the three-regime ordering. While the underlying preferences are private (making the game formally incomplete-information), the mechanism's design relies on dominant-strategy self-selection, which we argue approximates the complete-information outcome even when types are private. In the revision we will add an explicit discussion subsection clarifying this modeling choice, explaining why the SPNE provides a valid proxy for the welfare analysis, and including a brief argument (with supporting bounds) that the efficiency ordering and near-99% welfare results continue to hold approximately under private information. We will also detail how the simulation implements the strategy under private preferences. revision: partial

  2. Referee: [Numerical Evaluation] Numerical Evaluation section: The claim that the Price of Ignorance is near zero (>=99% welfare) rests on 5000 Monte Carlo runs across seven mechanisms. Without explicit details on preference distributions, parameter values, data-exclusion rules, or how the complete-information equilibrium is implemented under private preferences, it is unclear whether the result is robust or sensitive to post-hoc simulation choices, as the welfare gap is defined relative to an externally optimal benchmark.

    Authors: We agree that greater transparency on the simulation setup is necessary. The manuscript currently summarizes the Monte Carlo design at a high level; we will expand the Numerical Evaluation section (and add an appendix) with the precise details requested: the exact preference distributions (independent uniform draws on normalized [0,1] intervals for privacy and accuracy parameters), the full ranges of all other parameters, the data-exclusion criteria applied across the 5000 runs, and a step-by-step description of how the complete-information single-period selling strategy is coded and executed when preferences remain private to each user. We will also include additional sensitivity tables varying the distributions and parameter ranges to demonstrate that the >=99% welfare result is robust rather than an artifact of specific choices. revision: yes

Circularity Check

0 steps flagged

No circularity: derivation and numerical claim are independent of inputs

full rationale

The paper defines the Price of Ignorance explicitly as the welfare gap to an external personalized benchmark, proves a three-regime ordering under the complete-information SPNE assumption, and obtains the >=99% welfare result from separate Monte Carlo simulations (5000 runs, seven mechanisms). No equation or claim reduces the output to the input by construction, no parameter is fitted then relabeled as prediction, and no uniqueness theorem is imported via self-citation. The derivation chain remains self-contained against the stated model.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

Review based on abstract only; the central claim rests on standard game-theoretic rationality and equilibrium concepts plus simulation assumptions whose details are not provided.

axioms (1)
  • domain assumption Users are rational and play a subgame-perfect Nash equilibrium
    Invoked to establish the unique single-period selling equilibrium under complete information.

pith-pipeline@v0.9.0 · 5515 in / 1255 out tokens · 42270 ms · 2026-05-10T16:24:34.045062+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

47 extracted references · 47 canonical work pages

  1. [1]

    A survey on large language models for communication, network, and service management: Application insights, challenges, and future directions,

    G. O. Boateng, H. Sami, A. Alagha, H. Elmekki, A. Hammoud, R. Mizouni, A. Mourad, H. Otrok, J. Bentahar, S. Muhaidat, C. Talhi, Z. Dziong, and M. Guizani, “A survey on large language models for communication, network, and service management: Application insights, challenges, and future directions,”IEEE Communications Surveys & Tutorials, vol. 28, pp. 527–...

    work page 2026

  2. [2]

    General Data Protection Regulation,

    “General Data Protection Regulation,” 2016. [Online]. Available: https://gdpr-info.eu/

    work page 2016

  3. [3]

    California Consumer Privacy Act,

    “California Consumer Privacy Act,” 2018. [Online]. Available: https://oag.ca.gov/privacy/ccpa

    work page 2018

  4. [4]

    Personal Information Protection Law of the Peo- ple’s Republic of China,

    “Personal Information Protection Law of the Peo- ple’s Republic of China,” 2021. [Online]. Available: https://personalinformationprotectionlaw.com/

    work page 2021

  5. [5]

    Machine unlearning: A survey,

    H. Xu, T. Zhu, L. Zhang, W. Zhou, and P. S. Yu, “Machine unlearning: A survey,”ACM Comput. Surv., vol. 56, no. 1, pp. 1–36, Aug. 2023

    work page 2023

  6. [6]

    The price of forgetting: Data redemption mechanism design for machine unlearning,

    Y . Cui and M. H. Cheung, “The price of forgetting: Data redemption mechanism design for machine unlearning,” inIEEE ICC 2024, 2024, pp. 4054–4059

    work page 2024

  7. [7]

    A survey of data pricing for data marketplaces,

    M. Zhang, F. Beltr ´an, and J. Liu, “A survey of data pricing for data marketplaces,”IEEE Trans. Big Data, vol. 9, no. 4, pp. 1038–1056, 2023

    work page 2023

  8. [8]

    Price-discrimination game for distributed resource management in federated learning,

    H. Zhang, H. Yang, and G. Zhang, “Price-discrimination game for distributed resource management in federated learning,”IEEE Netw. Lett., vol. 6, no. 2, pp. 124–128, 2024

    work page 2024

  9. [9]

    Online pricing with reserve price constraint for personal data markets,

    C. Niu, Z. Zheng, F. Wu, S. Tang, and G. Chen, “Online pricing with reserve price constraint for personal data markets,”IEEE Trans. IEEE Trans. Knowl. Data Eng., vol. 34, no. 4, pp. 1928–1943, 2020

    work page 1928

  10. [10]

    The price of forgetting: Incentive mechanism design for machine unlearning,

    Y . Cui and M. H. Cheung, “The price of forgetting: Incentive mechanism design for machine unlearning,”IEEE Trans. Mobile Comput., vol. 24, no. 11, pp. 11 852–11 864, 2025

    work page 2025

  11. [11]

    Towards making systems forget with machine unlearning,

    Y . Cao and J. Yang, “Towards making systems forget with machine unlearning,” inProc. 36th IEEE Symp. Security and Privacy (S&P), 2015, pp. 463–480

    work page 2015

  12. [12]

    Machine unlearning,

    L. Bourtoule, V . Chandrasekaran, C. A. Choquette-Choo, H. Jia, A. Travers, B. Zhang, D. Lie, and N. Papernot, “Machine unlearning,” in Proc. 42nd IEEE Symp. Security and Privacy (S&P), 2021, pp. 141–159

    work page 2021

  13. [13]

    Making AI forget you: Data deletion in machine learning,

    A. Ginart, M. Guan, G. Valiant, and J. Y . Zou, “Making AI forget you: Data deletion in machine learning,” inAdvances in Neural Information Processing Systems 32 (NeurIPS), 2019, pp. 3513–3526

    work page 2019

  14. [14]

    Machine unlearning for random forests,

    J. Brophy and D. Lowd, “Machine unlearning for random forests,” in Proc. 38th Int. Conf. Machine Learning (ICML), ser. Proceedings of Machine Learning Research, vol. 139. PMLR, 2021, pp. 1092–1104

    work page 2021

  15. [15]

    Descent-to-delete: Gradient-based methods for machine unlearning,

    S. Neel, A. Roth, and S. Sharifi-Malvajerdi, “Descent-to-delete: Gradient-based methods for machine unlearning,” inProc. 32nd Int. Conf. Algorithmic Learning Theory (ALT), ser. Proceedings of Machine Learning Research, vol. 132. PMLR, 2021, pp. 931–962

    work page 2021

  16. [16]

    Markov chain Monte Carlo-based machine unlearning: Unlearn- ing what needs to be forgotten,

    Q. P. Nguyen, R. Oikawa, D. M. Divakaran, M. C. Chan, and B. K. H. Low, “Markov chain Monte Carlo-based machine unlearning: Unlearn- ing what needs to be forgotten,” inProceedings of the 2022 ACM on Asia Conference on Computer and Communications Security, 2022, pp. 351–363

    work page 2022

  17. [17]

    Adaptive machine unlearning,

    V . Gupta, C. Jung, S. Neel, A. Roth, S. Sharifi-Malvajerdi, and C. Waites, “Adaptive machine unlearning,” inAdvances in Neural Information Processing Systems 34 (NeurIPS), vol. 34, 2021, pp. 16 319–16 330

    work page 2021

  18. [18]

    Eternal sunshine of the spotless net: Selective forgetting in deep networks,

    A. Golatkar, A. Achille, and S. Soatto, “Eternal sunshine of the spotless net: Selective forgetting in deep networks,” inProc. IEEE/CVF Conf. Computer Vision and Pattern Recognition (CVPR), 2020, pp. 9301– 9309

    work page 2020

  19. [19]

    Remember what you want to forget: Algorithms for machine unlearning,

    A. Sekhari, J. Acharya, G. Kamath, and A. T. Suresh, “Remember what you want to forget: Algorithms for machine unlearning,” inAdvances in Neural Information Processing Systems 34 (NeurIPS), vol. 34, 2021, pp. 18 075–18 086

    work page 2021

  20. [20]

    Machine unlearning: Solutions and challenges,

    J. Xu, Z. Wu, C. Wang, and X. Jia, “Machine unlearning: Solutions and challenges,”IEEE Trans. Emerging Topics Comput. Intell., vol. 8, no. 3, pp. 2150–2168, 2024

    work page 2024

  21. [21]

    Right to be forgotten in the era of large language models: Implications, challenges, and solutions,

    D. Zhang, P. Finckenberg-Broman, T. Hoang, S. Pan, Z. Xing, M. Sta- ples, and X. Xu, “Right to be forgotten in the era of large language models: Implications, challenges, and solutions,”AI and Ethics, vol. 5, pp. 2445–2454, 2025

    work page 2025

  22. [22]

    FedEraser: Enabling efficient client-level data removal from federated learning models,

    G. Liu, X. Ma, Y . Yang, C. Wang, and J. Liu, “FedEraser: Enabling efficient client-level data removal from federated learning models,” in 2021 IEEE/ACM 29th Int. Symp. Quality of Service (IWQoS), 2021, pp. 1–10

    work page 2021

  23. [23]

    Hard to forget: Poisoning attacks on certified machine unlearning,

    N. G. Marchant, B. I. P. Rubinstein, and S. Alfeld, “Hard to forget: Poisoning attacks on certified machine unlearning,” inProc. 36th AAAI Conf. Artificial Intelligence, vol. 36, no. 7, 2022, pp. 7691–7700

    work page 2022

  24. [24]

    On the necessity of auditable algorithmic definitions for machine unlearning,

    A. Thudi, H. Jia, I. Shumailov, and N. Papernot, “On the necessity of auditable algorithmic definitions for machine unlearning,” inProc. 31st USENIX Security Symp., 2022, pp. 4007–4022

    work page 2022

  25. [25]

    A survey on data pricing: From economics to data science,

    J. Pei, “A survey on data pricing: From economics to data science,” IEEE Trans. Knowl. Data Eng., vol. 34, no. 10, pp. 4586–4608, 2022

    work page 2022

  26. [26]

    Data market platforms: Trading data assets to solve data problems,

    R. C. Fernandez, P. Subramaniam, and M. J. Franklin, “Data market platforms: Trading data assets to solve data problems,”Proc. VLDB Endowment, vol. 13, no. 12, pp. 1933–1947, 2020

    work page 1933

  27. [27]

    Too much data: Prices and inefficiencies in data markets,

    D. Acemoglu, A. Makhdoumi, A. Malekian, and A. Ozdaglar, “Too much data: Prices and inefficiencies in data markets,”American Eco- nomic Journal: Microeconomics, vol. 14, no. 4, pp. 218–256, 2022

    work page 2022

  28. [28]

    Query-based data pricing,

    P. Koutris, P. Upadhyaya, M. Balazinska, B. Howe, and D. Suciu, “Query-based data pricing,”Journal of the ACM, vol. 62, no. 5, pp. 1–44, 2015

    work page 2015

  29. [29]

    Data shapley: Equitable valuation of data for machine learning,

    A. Ghorbani and J. Zou, “Data shapley: Equitable valuation of data for machine learning,” inProc. 36th Int. Conf. Machine Learning (ICML), ser. Proceedings of Machine Learning Research, vol. 97. PMLR, 2019, pp. 2242–2251

    work page 2019

  30. [30]

    Towards efficient data valuation based on the shapley value,

    R. Jia, D. Dao, B. Wang, F. A. Hubis, N. Hynes, N. M. G ¨urel, B. Li, C. Zhang, D. Song, and C. J. Spanos, “Towards efficient data valuation based on the shapley value,” inProc. 22nd Int. Conf. Artificial Intelli- gence and Statistics (AISTATS), ser. Proceedings of Machine Learning Research, vol. 89. PMLR, 2019, pp. 1167–1176

    work page 2019

  31. [31]

    Beta shapley: A unified and noise-reduced data valuation framework for machine learning,

    Y . Kwon and J. Zou, “Beta shapley: A unified and noise-reduced data valuation framework for machine learning,” inProc. 25th Int. Conf. Artificial Intelligence and Statistics (AISTATS), ser. Proceedings of Machine Learning Research, vol. 151. PMLR, 2022, pp. 8780–8802

    work page 2022

  32. [32]

    A marketplace for data: An algorithmic solution,

    A. Agarwal, M. A. Dahleh, and T. Sarkar, “A marketplace for data: An algorithmic solution,” inProc. 2019 ACM Conf. Economics and Computation (EC), 2019, pp. 701–726

    work page 2019

  33. [33]

    Should we treat data as labor? Moving beyond “free

    I. Arrieta-Ibarra, L. Goff, D. Jim ´enez-Hern´andez, J. Lanier, and E. G. Weyl, “Should we treat data as labor? Moving beyond “free”,”AEA Papers and Proceedings, vol. 108, pp. 38–42, 2018. 17

    work page 2018

  34. [34]

    Data pricing in machine learning pipelines,

    Z. Cong, X. Luo, J. Pei, F. Zhu, and Y . Zhang, “Data pricing in machine learning pipelines,”Knowledge and Information Systems, vol. 64, pp. 1417–1455, 2022

    work page 2022

  35. [35]

    Incentive mech- anism for reliable federated learning: A joint optimization approach to combining reputation and contract theory,

    J. Kang, Z. Xiong, D. Niyato, S. Xie, and J. Zhang, “Incentive mech- anism for reliable federated learning: A joint optimization approach to combining reputation and contract theory,”IEEE Internet of Things J., vol. 6, no. 6, pp. 10 700–10 714, 2019

    work page 2019

  36. [36]

    A learning-based incentive mechanism for federated learning,

    Y . Zhan, P. Li, Z. Qu, D. Zeng, and S. Guo, “A learning-based incentive mechanism for federated learning,”IEEE Internet of Things J., vol. 7, no. 7, pp. 6360–6368, 2020

    work page 2020

  37. [37]

    An incentive mechanism for federated learning in wireless cellular networks: An auction approach,

    T. H. T. Le, N. H. Tran, Y . K. Tun, M. N. H. Nguyen, S. R. Pandey, Z. Han, and C. S. Hong, “An incentive mechanism for federated learning in wireless cellular networks: An auction approach,”IEEE Trans. Wireless Commun., vol. 20, no. 8, pp. 4874–4887, 2021

    work page 2021

  38. [38]

    Incentive mechanism for privacy-aware data aggregation in mobile crowd sensing systems,

    H. Jin, L. Su, H. Xiao, and K. Nahrstedt, “Incentive mechanism for privacy-aware data aggregation in mobile crowd sensing systems,” IEEE/ACM Trans. Networking, vol. 26, no. 5, pp. 2019–2032, 2018

    work page 2019

  39. [39]

    Differential privacy,

    C. Dwork, “Differential privacy,” inAutomata, Languages and Pro- gramming (ICALP), ser. Lecture Notes in Computer Science, vol. 4052. Springer, 2006, pp. 1–12

    work page 2006

  40. [40]

    Selling privacy at auction,

    A. Ghosh and A. Roth, “Selling privacy at auction,”Games and Economic Behavior, vol. 91, pp. 334–346, 2015

    work page 2015

  41. [41]

    The economics of privacy,

    A. Acquisti, C. R. Taylor, and L. Wagman, “The economics of privacy,” Journal of Economic Literature, vol. 54, no. 2, pp. 442–492, 2016

    work page 2016

  42. [42]

    Privacy and rationality in individual decision making,

    A. Acquisti and J. Grossklags, “Privacy and rationality in individual decision making,”IEEE Security & Privacy, vol. 3, no. 1, pp. 26–33, 2005

    work page 2005

  43. [43]

    Optimal auction design,

    R. B. Myerson, “Optimal auction design,”Mathematics of Operations Research, vol. 6, no. 1, pp. 58–73, 1981

    work page 1981

  44. [44]

    Existence and characterization of perfect equilibrium in games of perfect information,

    C. Harris, “Existence and characterization of perfect equilibrium in games of perfect information,”Econometrica, pp. 613–628, 1985

    work page 1985

  45. [45]

    Reexamination of the perfectness concept for equilibrium points in extensive games,

    R. Selten, “Reexamination of the perfectness concept for equilibrium points in extensive games,”International Journal of Game Theory, vol. 4, no. 1, pp. 25–55, 1975

    work page 1975

  46. [46]

    On the private provision of public goods,

    T. Bergstrom, L. Blume, and H. Varian, “On the private provision of public goods,”Journal of Public Economics, vol. 29, no. 1, pp. 25–49, 1986

    work page 1986

  47. [47]

    Why free ride? Strategies and learning in public goods experiments,

    J. Andreoni, “Why free ride? Strategies and learning in public goods experiments,”Journal of Public Economics, vol. 37, no. 3, pp. 291– 304, 1988

    work page 1988