The Public Option: A non-regulatory alternative to Network Neutrality - PowerPoint PPT Presentation

The Public Option: A non-regulatory alternative to Network Neutrality Richard Ma School of Computing National University of Singapore Joint work with Vishal Misra (Columbia University) The 2nd Workshop on Internet Economics

Highlights  A more realistic equilibrium model of content traffic, based on  User demand for content  System protocol/mechanism  Game theoretic analysis on user utility under different ISP market structures:  Monopoly, Duopoly & Oligopoly  Regulatory implications for all scenarios and the notion of a Public Option

Three-party model (𝑁, 𝜈, 𝒪) 𝝁 𝒋 𝝂 𝒪 𝑵 ⋮ ⋮  𝜈 : capacity of a single access ISP  𝑁 : # of users of the ISP (# of active users)  𝒪 : set of all content providers (CPs)  𝜇 𝑗 : throughput rate of CP 𝑗 ∈ 𝒪

User-side: 3 Demand Factors  Unconstrained throughput 𝜄 𝑗  Upper-bound, achieved under unlimited capacity  E.g. 5Mbps for Netflix  Popularity of the content 𝛽 𝑗  Google has a larger user base than other CPs.  Demand function of the content 𝑒 𝑗 (𝜄 𝑗 )  Percentage of users still being active under the achievable throughput 𝜄 𝑗 ≤ 𝜄 𝑗

Unconstrained Throughput 𝜇 𝑗 𝒋 (= 𝟖𝑳𝒄𝒒𝒕) User size 𝑵(= 𝟐𝟏) (Max) Throughput 𝜾 Content unconstrained throughput Content popularity 𝒋 = 𝜷 𝒋 𝑵𝜾 𝒋 (= 𝟓𝟑𝑳𝒄𝒒𝒕) 𝜷 𝒋 (= 𝟕𝟏%) 𝝁

Demand Function 𝒆 𝒋 𝜾 𝒋 demanding # of users 𝜷 𝒋 𝑵𝒆 𝒋 𝜾 𝒋 𝜷 𝒋 𝑵 achievable 𝒋 throughput 𝜾 𝜾 𝒋

Demand Function 𝒆 𝒋 𝜾 𝒋 demanding # of users 𝜷 𝒋 𝑵𝒆 𝒋 𝜾 𝒋  Assumption 1: 𝑒 𝑗 𝜄 𝑗 is continuous and non-decreasing in 𝜄 𝑗 with 𝑒 𝑗 𝜄 𝑗 = 1 .  More sensitive to throughput 𝜷 𝒋 𝑵  Throughput of CP i: 𝝁 𝒋 𝜾 𝒋 = 𝜷 𝒋 𝑵𝒆 𝒋 𝜾 𝒋 𝜾 𝒋 achievable 𝒋 throughput 𝜾 𝜾 𝒋

System Side: Rate Allocation  Axiom 1 (Throughput upper-bound) 𝑗 𝜄 𝑗 ≤ 𝜄  Axiom 2 (Work-conserving) 𝑗 𝜇 𝒪 = 𝜇 𝑗 = min 𝜈, 𝜇 𝑗∈𝒪 𝑗∈𝒪  Axiom 3 (Monotonicity) 𝜄 𝑗 𝑁, 𝜈 2 , 𝒪 ≥ 𝜄 𝑗 𝑁, 𝜈 1 , 𝒪 ∀ 𝜈 2 ≥ 𝜈 1

Uniqueness of Rate Equilibrium  Theorem (Uniqueness): A system (𝑁, 𝜈, 𝒪) has a unique equilibrium {𝜄 𝑗 ∶ 𝑗 ∈ 𝒪} (and therefore {𝜇 𝑗 ∶ 𝑗 ∈ 𝒪} ) under Assumption 1 and Axiom 1, 2 and 3. User demand: { 𝜄 𝑗 } → {𝑒 𝑗 } Rate allocation: μ, 𝑒 𝑗 → {𝜄 𝑗 }  Rate equalibrium: {𝜄 𝑗 ∗ }, {𝑒 𝑗 ∗ }

ISP Paid Prioritization ISP Payoff: 𝑑 = 𝑑𝜇 𝒬 𝜇 𝑗 𝑗∈𝒬 Capacity Charge Premium Class 𝝀𝝂 $ 𝒅 /unit traffic 𝑵, 𝝀𝝂, 𝓠 Ordinary Class (𝟐 − 𝝀)𝝂 $ 𝟏 𝑵, 𝟐 − 𝝀 𝝂, 𝓟

Monopolistic Analysis  Players: monopoly ISP 𝐽 and the set of CPs 𝒪  A Two-stage Game Model 𝑁, 𝜈, 𝒪, 𝐽  1 st stage, ISP chooses 𝑡 𝐽 = (𝜆, 𝑑) announces 𝑡 𝐽 .  2 nd stage, CPs simultaneously choose service classes reach a joint decision 𝑡 𝒪 = (𝒫, 𝒬) .  Outcome: set 𝒬 of CPs shares capacity 𝜆𝜈 and set 𝒫 of CPs share capacity 1 − 𝜆 𝜈 .

Utilities (Surplus)  ISP Surplus: 𝐽𝑇 = 𝑑 = 𝑑𝜇 𝒬 ; 𝜇 𝑗 𝑗∈𝒬  Consumer Surplus: 𝐷𝑇 = 𝜚 𝑗 𝜇 𝑗 𝑗∈𝒪  𝜚 𝑗 : per unit traffic value to the users  Content Provider:  𝑤 𝑗 : per unit traffic profit of CP 𝑗 𝑤 𝑗 𝜇 𝑗 if 𝑗 ∈ 𝒫, 𝑣 𝑗 𝜇 𝑗 = if 𝑗 ∈ 𝒬. 𝑤 𝑗 − 𝑑 𝜇 𝑗

Type of Content Profitability of CP 𝒘 𝒋 Value to users 𝝔 𝒋

Monopolistic Analysis  Players: monopoly ISP 𝐽 and the set of CPs 𝒪  A Two-stage Game Model 𝑁, 𝜈, 𝒪, 𝐽  1 st stage, ISP chooses 𝑡 𝐽 = (𝜆, 𝑑) announces 𝑡 𝐽 .  2 nd stage, CPs simultaneously choose service classes reach a joint decision 𝑡 𝒪 = (𝒫, 𝒬) .  Theorem: Given a fixed charge 𝑑 , strategy ′ = (1, 𝑑) . 𝑡 𝐽 = (𝜆, 𝑑) is dominated by 𝑡 𝐽  The monopoly ISP has incentive to allocate all capacity for the premium service class.

Utility Comparison: Φ vs 𝛺 𝜉 = 𝜈/𝑁 Φ = 𝐷𝑇/𝑁 Ψ = 𝐽𝑇/𝑁

Regulatory Implications  Ordinary service can be made “damaged goods”, which hurts the user utility.  Implication: ISP should not be allowed to use non-work-conserving policies ( 𝜆 cannot be too large).  Should we allow the ISP to charge an arbitrarily high price 𝑑 ?

High price 𝑑 is good when Profitability of CP 𝒘 𝒋 Value to users 𝝔 𝒋

High price 𝑑 is bad when Profitability of CP 𝒘 𝒋 Value to users 𝝔 𝒋

Oligopolistic Analysis  A Two-stage Game Model 𝑁, 𝜈, 𝒪, ℐ  1 st stage: for each ISP 𝐽 ∈ ℐ chooses 𝑡 𝐽 = (𝜆 𝐽 , 𝑑 𝐽 ) simultanously.  2 nd stage: at each ISP 𝐽 ∈ ℐ , CPs choose service 𝐽 = (𝒫 𝐽 , 𝒬 𝐽 ) classes with 𝑡 𝒪  Difference with monopolistic scenarios:  Users move among ISPs until the per user surplus Φ 𝐽 is the same, which determines the market share of the ISPs  ISPs try to maximize their market share.

Duopolistic Analysis 𝓠 ISP 𝑱 with 𝒕 𝑱 = (𝝀, 𝒅) 𝓟 ISP 𝑲 with 𝒕 𝑲 = (𝟏, 𝟏) 𝓞

Duopolistic Analysis: Results  Theorem: In the duopolistic game, where an ISP 𝐾 is a Public Option, i.e. 𝑡 𝐾 = (0, 0) , if 𝑡 𝐽 maximizes the non-neutral ISP 𝐽 ’s market share, 𝑡 𝐽 also maximizes user utility.  Regulatory implication for monopoly cases:

Oligopolistic Analysis: Results  Theorem: Under any strategy profile 𝑡 −𝐽 , if 𝑡 𝐽 is a best-response to 𝑡 −𝐽 that maximizes market share, then 𝑡 𝐽 is an 𝜗 – best-response for the per user utility Φ .  The Nash equilibrium of market share is an 𝜗 -Nash equilibrium of user utility.  Oligopolistic scenarios:

Regulatory Preference ISP market structure Oligopoly Monopoly User Utility