This book discusses the pros and cons of information and communication (ICT) neutrality. It tries to be as objective as possible from arguments of proponents and opponents, this way enabling readers to build their own opinion. It presents the history of the ongoing network neutrality debate, the various concepts it encompasses, and also some mathematical developments illustrating optimal strategies and potential counter-intuitive results, then extends the discussion to connected ICT domains. The book thus touches issues related to history, economics, law, networking, and mathematics.
After an introductory chapter on the history of the topic, chapter 2 surveys and compares the various laws in place worldwide and discusses some implications of heterogeneous rules in several regions. Next, chapter 3 details the arguments put forward by the participants of the net neutrality debate. Chapter 4 then presents how the impact of neutral or non-neutral behaviors can be analyzed mathematically, with sometimes counter-intuitive results, and emphasizes the interest of modeling to avoid bad decisions. Chapter 5 illustrates that content providers may not always be on the pro-neutrality side, as there are situations where they may have an economic advantage with a non-neutral situation, e.g. when they are leaders on a market and create barriers to entry for competitors. Another related issue is covered in chapter 6, which discusses existing ways for ISPs to circumvent the packet-based rules and behave non-neutral without breaking the written law. Chapter 7 gives more insight on the role and possible non-neutral behavior of search engines, leading to another debate called the search neutrality debate. Chapter 8 focuses on e-commerce platforms and social networks, and investigates how they can influence users’ actions and opinions. The issue is linked to the debate on the transparency of algorithms which is active in Europe especially. Chapter 9 focuses on enforcing neutrality in practice through measurements: indeed, setting rules requires monitoring the activity of ICT actors in order to sanction non-appropriate behaviors and be proactive against new conducts. The chapter explains why this is challenging and what tools are currently available. Eventually, Chapter 10 briefly concludes the presentation and opens the debate.
Author(s): Patrick Maillé, Bruno Tuffin
Publisher: Springer
Year: 2022
Language: English
Pages: 193
City: Cham
Preface
Acknowledgments
Contents
About the Authors
Acronyms
1 Introduction: A Bit of History
1.1 The Advent of the Digital Economy and the Need for Regulation
1.2 The Internet: Worries About the Lack of Neutrality
1.3 New Services: New Difficulties and Potential Hindrance to Society
1.4 Toward ICT Neutrality?
1.5 Goal and Organization of This Book
2 Definitions
2.1 Several Definitions of Neutrality
2.2 Analysis and Comparison
2.3 Specialized Services
2.3.1 Perimeter of Specialized Services
2.3.2 Need for Specialized Services
2.3.3 Implications of Introducing a Special Class of Service
2.4 Zero-Rating and Sponsored Data
2.5 Public Consultations
2.6 Worldwide Implementations
2.6.1 North America
2.6.1.1 USA
2.6.1.2 Canada
2.6.1.3 Mexico
2.6.2 Europe
2.6.2.1 European Union
2.6.2.2 France
2.6.2.3 Germany
2.6.2.4 United Kingdom
2.6.2.5 The Netherlands
2.6.2.6 Norway
2.6.2.7 Slovenia
2.6.2.8 Russia
2.6.3 Asia
2.6.3.1 China
2.6.3.2 Japan
2.6.3.3 South Korea
2.6.3.4 India
2.6.4 Oceania
2.6.4.1 Australia
2.6.4.2 New Zealand
2.6.5 South America
2.6.5.1 Chile
2.6.5.2 Brazil
2.6.6 Africa
2.7 A Situation Still Evolving
2.8 A Trade-Off with Users Deciding Differentiation?
2.9 Conclusions
3 Pros and Cons: The Arguments in the Debate
3.1 The ``Freedom'' Arguments: Whose Freedom?
3.1.1 [Pro-NN] Net Neutrality Is Needed to Guarantee Free Speech
3.1.2 [Anti-NN] Net Neutrality Hinders Operators' Freedom of Enterprise
3.2 The ``Investment'' Arguments
3.2.1 [Anti-NN] Returns on Investments Are Needed
3.2.2 [Pro-NN] More Customers, More Investments
3.3 The ``Competition and Innovation'' Arguments
3.3.1 [Pro-NN] Non-neutrality Biases the Competition Among Services
3.3.2 [Anti-NN] ISPs Favoring Some Content Providers Stimulates ISP Competition (Through Content)
3.3.3 [Pro-NN] Neutrality Is Needed to Allow Innovating Newcomers
3.4 The ``Security and Privacy'' Arguments
3.4.1 [Anti-NN] Encryption Already Prevents Discrimination
3.4.2 [Pro-NN] Neutrality Helps Protect Privacy
3.5 The ``Welfare'' Arguments
3.5.1 [Pro-NN] Non-neutrality Would Lead to a Fragmented Internet, Which Is Bad For Users
3.5.2 [Anti-NN] Content Providers Should Contribute to the Infrastructure Costs to Reduce Users' Share
3.5.3 [Anti-NN] A Free Market Finds the Most Efficient Outcome
3.6 Conclusions
4 Mathematical Analysis
4.1 Introduction to Mathematical Modeling and Game Theory
4.1.1 Elements of Non-cooperative Game Theory
4.1.2 Output Analysis
4.1.3 Principle of Mechanism Design
4.2 A Basic Model
4.3 Model with Two Content Providers
4.3.1 Pricing Game Between CPs (α=2)
4.3.2 Fully Neutral Case (α=2)
4.3.3 Weakly Neutral Case (α=2)
4.3.4 Non-neutral Case (α=2)
4.3.5 Comparison
4.4 Other Conclusions from the Literature
4.5 Additional Notes
5 Non-neutrality Pushed by Content Providers
5.1 Non-neutrality as a Barrier to Entry Protecting Incumbents
5.1.1 Modeling the Impact of a New Entrant CP
5.1.2 Actors' Revenues and Decision Time Scales
5.1.2.1 CP Revenues
5.1.2.2 ISP Revenue
5.1.2.3 Order of Decisions
5.1.3 Analyzing the Game for Fixed Side Payments
5.1.3.1 User Decisions
5.1.3.2 ISP Subscription Price pA
5.1.3.3 Should the New CP Enter the Market?
5.1.3.4 Illustration Example: Impact of the Side Payment
5.1.4 Can the ISP and the Incumbent CP Agree on Side Payments?
5.1.4.1 Side Payments Maximizing the ISP Revenue Are Not Sustainable
5.1.4.2 Can Side Payments Benefit Both the Incumbent CP and the ISP?
5.1.4.3 How Can the ISP and Incumbent CP Agree on a Side Payment Level?
5.1.4.4 What About Users? Should the Regulator Intervene?
5.1.5 Refining the Model
5.2 Sponsored Data and Zero Rating
5.2.1 Modeling Sponsored-Data Practices in a Competitive ISP Context
5.2.1.1 Users' Decisions and Preferences
5.2.1.2 The CP Side: Advertising and Sponsoring
5.2.1.3 ISP Price Competition
5.2.1.4 Order of Decisions
5.2.2 How Does the CP Decide to Sponsor and Add Advertisement?
5.2.2.1 The Pricing Game Played by ISPs
5.2.3 A Practice Finally Mostly Benefiting ISPs?
5.3 When Big CPs Designate the ``Good'' ISPs: Incentivizing Non-neutrality Without Paying
5.3.1 ISP Competition: The Importance of Looking Good
5.3.2 A Two-ISP, Two-CP model with Heterogeneous Users
5.3.2.1 Modelling Consumer (Heterogeneous) Usage of Different CPs
5.3.2.2 ISPs Deciding How to Allocate Their Capacity Affects CP Quality
5.3.2.3 Introducing Some Asymmetry Among CPs: Required Throughput
5.3.2.4 Users Care About the Quality of the Services They Use!
5.3.2.5 A (Simple) Game on Differentiation
5.3.3 User Equilibria: How ISP Decisions Affect User Subscription Choices
5.3.4 The Game Among ISPs and Its Consequences
5.3.4.1 At the Equilibrium, Both ISPs Differentiate the Same Way
5.3.4.2 Equilibrium vs Neutral Capacity Allocation: Neutrality Hindering Innovation?
5.3.4.3 Equilibrium vs Neutral Capacity Allocation: No Impact for ISPs
5.3.4.4 Equilibrium vs Neutral Capacity Allocation: What About Users?
5.3.5 A Model Yielding Unexpected Conclusions
5.4 Conclusions
6 A More General View of Neutrality
6.1 Is It Relevant to Generalize Network Neutrality?
6.2 Content Delivery Networks: Intermediaries Flying Under the Radar?
6.2.1 Introduction to CDN Role and Impact
6.2.2 Model Illustrating the Impact of CDNs on Competition Between CPs
6.2.3 Model Illustrating the Impact of CDNs on Competition Between ISPs
6.3 Issues Related to Vertical Integration
6.3.1 Vertical Integration and the Internet Ecosystem
6.3.2 Illustration of Vertical Integration on Our Basic Model
6.3.3 Other Illustrations and Results from Models
6.4 Device Neutrality
6.5 Neutrality of Structuring Platforms
6.6 Also Toward Data Neutrality?
7 Search Neutrality
7.1 Is It in Search Engines' Interest to Be Neutral?
7.1.1 Modeling the Long-Term vs. Short-Term Gains Trade-Off
7.1.1.1 Modeling Requests as Random Pairs of Vectors
7.1.1.2 Click-Through-Rate: Impact of Position and Relevance
7.1.1.3 Frequency of Requests and SE Objective Function
7.1.1.4 How Would a Neutral SE Work?
7.1.2 The Revenue-Maximizing Ranking Policy: An (Almost) Explicit Trade-Off Between Relevance and Gains
7.1.3 Neutral vs. Non-neutral Search Engine: What Are the Differences?
7.1.3.1 A Simple Example with 10 Actors
7.1.3.2 Ad Revenues Are Good for Result Quality!
7.1.3.3 Should Users Care Whether Their SE Is Neutral?
7.1.3.4 Quantifying How Non-neutrality Distorts Competition Among CPs
7.1.3.5 Impact of Search (Non-)Neutrality on Innovation
7.2 Detecting Bias and Trying to Build a ``Neutral'' Search Engine
7.2.1 A Look At Related Contributions
7.2.2 An Estimation of Page Relevance: Average Page Visibility
7.2.3 Quantifying How ``Consensual'' a Search Engine Is:The SE Score
7.2.4 Pointing Out Possible Bias: Statistical Tests
7.2.5 Aggregating Results from SEs to Create Meta-SEs
7.2.5.1 The Consensus Ranking
7.2.5.2 The Majority Judgment Ranking
7.2.5.3 Robustness to Bias: Consensus Ranking vs Majority Judgment
7.2.6 A Few Observations from a Campaign on 1000 Search Terms
7.2.6.1 The ``Chrome'' Search Term
7.3 Conclusion
8 Algorithmic Transparency
8.1 Algorithmic Transparency and Structuring Platforms: Why Introduce Scrutiny?
8.2 Worries About the Transparency of Artificial Intelligence Algorithms
8.3 How to Proceed?
8.4 Fake News, Manipulation of Opinions, Hatred, and Fact Checking
8.5 Conclusion on Transparency and Neutrality
9 Tools to Monitor Neutrality
9.1 Why Are Specific Tools Needed?
9.2 Tools for the Monitoring of ISPs
9.2.1 Potential Infringements
9.2.2 Available Tools
9.2.3 The Limits of the Existing Tools
9.3 Monitoring Other Actors
9.3.1 Monitoring CDNs
9.3.2 Monitoring Search Engines
9.4 Tools and Algorithmic Transparency
9.5 Fact-Checking Tools
10 Conclusions
References