• Skip to main content
  • Skip to secondary menu
  • Skip to primary sidebar
  • Skip to footer
Montreal AI Ethics Institute

Montreal AI Ethics Institute

Democratizing AI ethics literacy

  • Articles
    • Public Policy
    • Privacy & Security
    • Human Rights
      • Ethics
      • JEDI (Justice, Equity, Diversity, Inclusion
    • Climate
    • Design
      • Emerging Technology
    • Application & Adoption
      • Health
      • Education
      • Government
        • Military
        • Public Works
      • Labour
    • Arts & Culture
      • Film & TV
      • Music
      • Pop Culture
      • Digital Art
  • Columns
    • AI Policy Corner
    • Recess
    • Tech Futures
  • The AI Ethics Brief
  • AI Literacy
    • Research Summaries
    • AI Ethics Living Dictionary
    • Learning Community
  • The State of AI Ethics Report
    • State of AI Ethics Report Volume 8 (2026): Call for Contributors
    • Volume 7 (November 2025)
    • Volume 6 (February 2022)
    • Volume 5 (July 2021)
    • Volume 4 (April 2021)
    • Volume 3 (Jan 2021)
    • Volume 2 (Oct 2020)
    • Volume 1 (June 2020)
  • About
    • Our Contributions Policy
    • Our Open Access Policy
    • Contact
    • Donate

Regulatory Instruments for Fair Personalized Pricing

March 30, 2022

šŸ”¬ Research Summary by Renzhe Xu, a third year PhD student in Computer Science and Technology at Tsinghua University and his research interests include fair machine learning, causal inference, and data mining.

[Original paper by Renzhe Xu, Xingxuan Zhang, Peng Cui, Bo Li, Zheyan Shen, Jiazheng Xu]


Overview: Ā How can we regulate personalized pricing to ensure consumers’ benefits? In this work, we propose two simple but effective regulatory policies under an idealized circumstance. Our findings and insights shed light on regulatory policy design for the increasingly monopolized business in the digital era.


Introduction

Personalized pricing is a business strategy to charge different prices to individual consumers based on their characteristics and behaviors. It has become common practice in many industries nowadays due to the availability of a growing amount of high granular consumer data. The discriminatory nature of personalized pricing has triggered heated debates among policymakers and academics on how to design regulation policies to balance market efficiency and equity. In this work, we propose two sound policy instruments and investigate the optimal pricing strategy of a profit-maximizing monopoly under both regulatory constraints. We analyze the impact of imposing the proposed policies on consumer surplus, producer surplus, and social welfare. We show that both proposed constraints can help balance consumer surplus and producer surplus at the expense of total surplus for common demand distributions.

Background

Personalized pricing, once considered the idealized construction of economic theories, has become common practice in many industries due to the availability of the increasing amount of consumer data. The main concern of personalized pricing is that it transfers value from consumers to shareholders, increasing inequality and inefficiency from a utilitarian standpoint. As a result, effective regulatory policies are required to balance the benefits between consumers and companies, which are measured as consumer surplus and producer surplus, respectively.

In this work, we study in designing effective policy instruments to balance consumer surplus and producer surplus. We consider the most straightforward scenario where a monopoly sells a single product with fixed marginal cost to different consumers. In addition, we assume that the monopoly can precisely estimate each consumers’ willingness to pay and the purpose of the monopoly is to find a personalized pricing strategy to maximize its revenue while remaining compliant with the policy instruments.

Proposed Policy instruments

We propose two sound policy instruments and prove their effectiveness in balancing consumer surplus and producer surplus. These two policies, named the difference and ratio constraints, are introduced to regulate the range of personalized prices by constraining the difference and ratio between the maximal price and minimal price, respectively. For example, a $2-difference constraint allows a pricing range [$1, $3] (3 – 1 = 2) but disallows [$1, $4] (4 – 1 = 3 > 2) while, a 2-ratio constraint allows [$2, $4] (4 / 2 = 2) but disallows [$2, $5] (5 / 2 = 2.5 > 2).

The constant in the constraints (such as $2 in the $2-difference constraint shown above) determines the intensity of the regulation. A smaller constant will lead to a stricter regulatory intensity in both constraints. Now consider the most extreme cases. On the one hand, when the constant becomes 0 in the difference constraint or becomes 1 in the ratio constraint, the regulatory intensity is the most strict and only uniform pricing (the price is fixed for all consumers) is allowed. On the other hand, when the constant becomes big enough, the regulation is very slight and perfect price discrimination is possible.

Theoretical results on consumer surplus, producer surplus, and social welfare

The results on social welfare analysis under both constraints are threefold. Graphical explanations of these findings are shown in the figure.

1. Both constraints can effectively balance the consumer surplus and producer surplus, which means the consumer surplus increases while the producer surplus decreases as the constraints become stricter.

2. We compare the trade-off between consumer surplus and producer surplus achieved by the two constraints and show that the difference constraint outperforms the ratio constraint. This means that the consumer surplus under the ratio constraint is smaller than that under the difference constraint if the producer surplus under the two constraints is equal.

3. Imposing either of the constraints will inevitably, to some extent, harm the total surplus. This result is reasonable, given that the efficiency-equity trade-off is largely recognized in practice. In addition, the perfect price discrimination achieves the maximal market efficiency and any regulatory policies to avoid it will inevitably harm the total surplus.

Empirical results

We run experiments on several real-world datasets to verify our theoretical results. The details of the datasets are provided as follows.

1. Coke and cake. Wertenbroch and Skiera [2002] adopted Becker, DeGroot, and Marschak’s method to estimate willingness-to-pay for a can of Coca-Cola on a public beach and a piece of pound cake on a commuter ferry in Kiel, Germany.

2. Elective vaccine. Slunge [2015] studied willingness to pay for vaccination against tick-borne encephalitis in Sweden. They asked individuals with personal information about take-up at a random price of 100, 250, 500, 750, or 1000 SEK.

3. Auto loan. The dataset records 208,085 auto loan applications received by a major online lender in the United States with loan-specific features. Following prior works, we adopt the feature selection results and consider only four features. The price of a loan is computed as the net present value of future payment minus the loan amount.

The trade-off curves between consumer surplus and producer surplus under our proposed constraints in these datasets are shown in the figure. The figure verifies that both constraints can help balance consumer surplus and producer surplus at the expense of total surplus. In addition, the curve of the ratio constraint is on top of the difference constraint.

Between the lines

We should notice that it is important to analyze the impacts of policies in fair personalized pricing theoretically. Improperly designed policies may harm both consumers and producers. For example, Dubé and Misra [2021] found that finer-grained personalized pricing in third-degree price discrimination can increase consumer welfare, which is contrary to the common belief that personalized pricing will always harm consumers.

In addition, comparatively, the ratio constraint is more suitable for designing policies because ratios could be easily adapted to various scenarios. The difference constraint has better performance on the trade-off between consumer surplus and producer surplus. As a result, the two constraints could be adopted in different applications in practice.

Want quick summaries of the latest research & reporting in AI ethics delivered to your inbox? Subscribe to the AI Ethics Brief. We publish bi-weekly.

Primary Sidebar

SAIER Volume 8 (2026)

SAIER Volume 8 (2026) Call for Contributors

šŸ” SEARCH

Spotlight

Tech Futures: Introducing the Resist List

An abstract spiral of dark circles appears at the centre, resembling a tornado. Several vintage magazine covers and advertisements are being drawn toward the spiral. The artworks that have already been pulled into it are becoming distorted and replaced with clusters of numbers representing their numerical embeddings.

Tech Futures: Better Imagination for Better Tech Futures

This image is a collage with a colourful Japanese vintage landscape showing a mountain, hills, flowers and other plants and a small stream. There are 3 large black data servers placed in the bottom half of the image, with a cloud of black smoke emitting from them, partly obscuring the scenery.

Tech Futures: Crafting Participatory Tech Futures

A network diagram with lots of little emojis, organised in clusters.

Tech Futures: AI For and Against Knowledge

A brightly coloured illustration which can be viewed in any direction. It has many elements to it working together: men in suits around a table, someone in a data centre, big hands controlling the scenes and holding a phone, people in a production line. Motifs such as network diagrams and melting emojis are placed throughout the busy vignettes.

Tech Futures: The Fossil Fuels Playbook for Big Tech: Part II

related posts

  • Achieving a ā€˜Good AI Society’: Comparing the Aims and Progress of the EU and the US

    Achieving a ā€˜Good AI Society’: Comparing the Aims and Progress of the EU and the US

  • Evaluating a Methodology for Increasing AI Transparency: A Case Study

    Evaluating a Methodology for Increasing AI Transparency: A Case Study

  • Research summary: AI Governance in 2019, A Year in Review: Observations of 50 Global Experts

    Research summary: AI Governance in 2019, A Year in Review: Observations of 50 Global Experts

  • AI Consent Futures: A Case Study on Voice Data Collection with Clinicians

    AI Consent Futures: A Case Study on Voice Data Collection with Clinicians

  • Seeing Like a Toolkit: How Toolkits Envision the Work of AI Ethics

    Seeing Like a Toolkit: How Toolkits Envision the Work of AI Ethics

  • Research Summary: Geo-indistinguishability: Differential privacy for location-based systems

    Research Summary: Geo-indistinguishability: Differential privacy for location-based systems

  • AI vs. Maya Angelou: Experimental Evidence That People Cannot Differentiate AI-Generated From Human-...

    AI vs. Maya Angelou: Experimental Evidence That People Cannot Differentiate AI-Generated From Human-...

  • A Taxonomy of Foundation Model based Systems for Responsible-AI-by-Design

    A Taxonomy of Foundation Model based Systems for Responsible-AI-by-Design

  • Risky Analysis: Assessing and Improving AI Governance Tools

    Risky Analysis: Assessing and Improving AI Governance Tools

  • Moral Machine or Tyranny of the Majority?

    Moral Machine or Tyranny of the Majority?

Partners

  • Ā 
    U.S. Artificial Intelligence Safety Institute Consortium (AISIC) at NIST

  • Partnership on AI

  • The LF AI & Data Foundation

  • The AI Alliance

Footer


Articles

Columns

AI Literacy

The State of AI Ethics Report


 

About Us


Founded in 2018, the Montreal AI Ethics Institute (MAIEI) is an international non-profit organization equipping citizens concerned about artificial intelligence and its impact on society to take action.

Contact

Donate


  • Ā© 2025 MONTREAL AI ETHICS INSTITUTE.
  • This work is licensed under a Creative Commons Attribution 4.0 International License.
  • Learn more about our open access policy here.
  • Creative Commons License

    Save hours of work and stay on top of Responsible AI research and reporting with our bi-weekly email newsletter.