Information Asymmetry in Queues with Strategic Customers

Basic Information

• Paper ID: 2510.09899
• Title: Information Asymmetry in Queues with Strategic Customers
• Authors: Shunan Zheng, John Hasenbein (University of Texas at Austin)
• Classification: math.OC (Optimization and Control)
• Publication Date: October 14, 2025
• Paper Link: https://arxiv.org/abs/2510.09899

Abstract

This paper investigates information asymmetry in unobservable single-server queueing systems. While system managers possess knowledge of the true arrival rate, customers may lack this information and form arbitrary beliefs. The paper proposes a three-tier information asymmetry hierarchy with increasing levels of information disclosure: customers maintain private beliefs, customers are aware of others' beliefs, and customers know the true arrival rate. Within this framework, the paper analyzes how general belief distributions affect equilibrium joining probabilities, revenues, and social welfare, and proposes information disclosure strategies for system managers to regulate queues.

Research Background and Motivation

Research Questions

Traditional strategic queueing theory assumes customers possess complete system information. However, in reality, customers typically have more limited information than managers, and information levels depend heavily on what managers choose to disclose. This information asymmetry creates uncertainty for customers, who must make decisions based on incomplete information.

Importance of the Problem

1. Practical Relevance: Third-party platforms have recognized the importance of narrowing information gaps between system managers and customers, typically providing historical congestion estimates and real-time system status updates
2. Decision Impact: Information asymmetry directly affects customers' joining strategies, which in turn impacts system revenue and social welfare
3. Managerial Significance: Provides managers with guidance on joint decisions regarding pricing and information disclosure

Limitations of Existing Approaches

Existing literature often assumes simple two-class belief structures, whereas this paper studies general forms of belief distributions, making conclusions more applicable to various scenarios. Additionally, prior research has given limited systematic attention to information disclosure as a regulatory tool.

Core Contributions

1. Proposes a Three-Tier Information Asymmetry Framework: Private beliefs, shared beliefs, and the classical case (complete information)
2. Analyzes the Impact of General Belief Distributions: Systematic analysis regarding equilibrium joining probabilities, revenues, and social welfare
3. Establishes Threshold Theory for Revenue Comparison: Provides precise conditions for revenue relationships under different information levels
4. Provides Management Decision Guidance: Offers information disclosure strategies for revenue maximizers (RM) and social optimizers (SO)

Methodology Details

Task Definition

Studies information asymmetry in M/G/1 queueing systems where:

• Inputs: Arrival rate λ, service rate μ, customer belief distribution Λ, system parameters (R, C, p)
• Outputs: Equilibrium joining probability, optimal revenue, social welfare
• Constraints: ρ = λ/μ < 1, R ≥ C/μ

Model Architecture

Three-Tier Information Structure

1. Classical Case: Customers know the true arrival rate λ; all customers adopt the same strategy q
2. Shared Belief Case: Customers are aware of belief distribution Λ and calculate utility based on expectations
3. Private Belief Case: Each customer makes decisions independently based on personal belief λ̃

Mathematical Model

Customer utility function:

U = R - p - CW(λe)

where W(λe) is the expected waiting time and λe is the effective arrival rate.

Revenue function (private belief case):

RevP(p) = pλE[Q(p)]

where Q(p) = min{ξ(p)/Λ, 1} and ξ(p) is the threshold effective arrival rate.

Social welfare function:

SWP(p) = E[λQ(p)(R - CW(λQ(p)))]

Technical Innovations

1. Threshold Function M(ξ): Establishes precise thresholds for comparing private belief and classical case revenues
2. General Belief Distribution Treatment: Does not restrict belief distribution form; only requires bounded support
3. Information Value Quantification: Systematically compares economic value across different information levels

Experimental Setup

Parameter Configuration

• System parameters: R = 5, C = 4 (or 5), μ = 4 (or 5)
• True arrival rate: λ = 3 (or 4.2)
• Belief distributions: Uniform distribution U(a,b), two-point discrete distribution

Evaluation Metrics

• Equilibrium joining probability (q)
• Revenue rate (Rev)
• Social welfare rate (SW)
• Threshold function M(ξ)

Comparison Methods

Systematic comparison across three information levels:

• Classical case vs. shared belief
• Private belief vs. classical case
• Private belief vs. shared belief

Experimental Results

Main Results

Equilibrium Relationships

Under unbiased waiting time conditions:

qSm ≤ qCs = qCm = qSs ≤ qCe = qSe

Revenue Comparison Thresholds

1. Private Belief vs. Classical Case: When λ > M(ξ), RevP > RevC
2. Shared Belief vs. Classical Case: When λ ≤ EΛ, RevS ≤ RevC
3. Optimal Revenue: When λ ≤ 1/E1/Λ, RevPopt ≤ RevCopt

Numerical Verification

Experiment 1 shows that as the belief distribution mean varies from 3.7 to 4.7, M(ξ) increases from 3.719 to 4.694, validating the accuracy of theoretical thresholds.

Experiment 2 demonstrates that increasing the variance of the belief distribution improves revenue in the pessimistic customer case, but may reduce revenue when customers are overly optimistic.

Management Strategy Guidance

• RM Strategy: Disclose true arrival rate when λ ≤ 1/E1/Λ; otherwise, calculate M(ξC) for comparison with λ
• SO Strategy: Always disclose information when customers are optimistic; still disclose when pessimistic but within specific ranges

Related Work

Classical Theoretical Foundations

• Naor (1969) first studied strategic customer behavior in queues
• Edelson & Hildebrand (1975) studied unobservable queues

Research on Parameter Uncertainty

• Hassin & Snitkovsky (2020): "Naor inequality" under arrival rate uncertainty
• Chen & Hasenbein (2020): Objectives of social optimizers and revenue maximizers diverge under arrival rate uncertainty
• Cui & Veeraraghavan (2016): Impact of arbitrary service rate beliefs in observable queues

Advantages Relative to Existing Work

1. Studies general forms of belief distributions rather than simple two-class structures
2. Systematic three-tier information framework
3. Provides precise threshold conditions and management decision guidance

Conclusions and Discussion

Main Conclusions

1. Information asymmetry significantly affects equilibrium behavior, following specific partial order relationships
2. Revenue impact depends on the relationship between belief distribution and true parameters, with precise thresholds existing
3. Social welfare is typically optimal under complete information, with pessimistic customers as exceptions
4. Managers should develop differentiated information disclosure strategies based on customer belief characteristics

Limitations

1. Assumes managers reset prices daily, which may be unrealistic in practice
2. Does not consider customer learning and belief updating
3. Focuses only on arrival rate uncertainty; uncertainty in other parameters remains to be studied

Future Directions

1. Cases where managers also face uncertainty
2. Belief updating mechanisms for Bayesian customers
3. Extension to uncertainty in other system parameters

In-Depth Evaluation

Strengths

1. Theoretical Rigor: Provides a complete mathematical analysis framework with detailed proofs
2. Practical Value: Offers actionable decision guidance for actual queueing system management
3. Strong Generality: Applicable to arbitrary belief distributions without restricting specific forms
4. Systematic Coverage: Three-tier framework comprehensively covers various information asymmetry scenarios

Weaknesses

1. Strong Model Assumptions: Assumptions such as non-learning customers and complete managerial information may limit applicability
2. High Complexity: Some results require numerical computation; lacks closed-form solutions
3. Insufficient Empirical Validation: Primarily theoretical analysis with limited real-world data verification

Impact

This research provides important theoretical contributions to the intersection of information economics and queueing theory, with significant implications for service industry management, particularly regarding information transparency management in the digital platform era.

Applicable Scenarios

• Information disclosure strategies for online service platforms regarding congestion
• Appointment system management in medical services
• Information dissemination strategies in transportation systems
• Waiting time management in food service industries

References

1. Naor, P. (1969). The Regulation of Queue Size by Levying Tolls. Econometrica.
2. Edelson, N. M., & Hilderbrand, D. K. (1975). Congestion Tolls for Poisson Queuing Processes. Econometrica.
3. Hassin, R. (2016). Rational Queueing. Chapman and Hall/CRC.
4. Cui, S., & Veeraraghavan, S. (2016). Blind Queues: The Impact of Consumer Beliefs on Revenues and Congestion. Management Science.