Network Neutrality in the Developing World Panel

From A2K Wiki

Notes on the panel

Mike Steffen:

• Panel history:
  • Eddan Katz: Let's talk about Net Neutrality in the Developing World.
  • Mike: I don't think much has written about that.
  • Eddan Katz: Exactly.
• Mike Godwin has been leading the field in this area.

Mike Godwin:

• To what extent is Network Neutrality a "missing piece" of the A2K spectrum? It's so often about intellectual property, or knowledge policy in the developing world.
• Book "Technologies of Freedom" by Pool: Two notions in US system that protect freedom. The first is government non-discrimination as to content; the second is common carriage - communications carriers do not discriminate as to content. Today, he would have added a third - network neutrality.
  • Neutrality as to applications - free to develop and run any type of application that leads to the creation and dissemination of knowledge.
• There's no settled paradigm about the meaning of network neutrality and about what the paradigm ought to be.
  • Division exists in policy perspectives on this issue:
    • Those who believe that we are all better off with an open, neutral network
    • Those who believe that we need to protect incentives to service providers to innovate, and that carriers should be able to discriminate as to content.

Michael Geist:

• Law Professor at University of Ottawa

• Title of talk: The Two-Tier Internet

• Not fully equipped to discuss developing world (from Canada), but some of the same issues are arising in all jurisdictions.
• Starting point: Are all bits bits? Should they all be treated equally?
  • Lessig et al.: End-to-end network is best; no carrier discrimination.
  • Concern of limited carrier options - e.g. in Canada, only cable or DSL.
• Other developing problems
  • Developing technologies such as VoIP are introducing lots of competition
    • Example: Some ISPs imposing a surcharge for the use of VoIP.
    • Many countries, such as Panama and Egypt, are blocking VoIP to maintain a monopoly on telecommunications services.
  • Other forms of content have also been blocked in Canada.
    • Telus customers blocked access to Voices for Change website in July 2005.
    • Telus blocked access first, and asked for a court ordered response later.
    • Telecomm lawyers did not know of any way to stop them.
  • Risk of traffic shaping - prioritizing some content over others
    • In particular, file sharing and podcasting are given very low priorities
    • There is no notice of priority policies.
  • Private v. Public Internet
    • E.g. Verizon's separate service for HDTV and web content
    • ISPs are making private deals with content providers to limit access to the high speed services
  • Website "premiums"
    • Charge websites to enable high-speed access by visitors
• Policy questions
  • Do we need legal protection for or from internet tiering?
  • Is tiering needed for proper incentives?
  • General A2K North/South questions
• Legal protections?
  • Tiering does exist today (broadband, dial-up, wireless, etc.).
    • But access tiering (per-site basis) is a different problem - it introduces competition concerns.
  • Ultimately, there has been a call for change in Canadian law to enforce net neutrality against access tiering.
• Incentives
  • There is plenty of rhetoric but little evidence supporting access tiering.
  • High cost to enabling the distinctions, but little reward
• North/South
  • Eventually South may be even more limited, given reliance on VoIP.

Susan Crawford:

• Law Professor at Cardozo

• Title: Seeing the Net: U.S. Telecomm Policy

• General commentary
  • The two sides (regulation or not) pass in the night; little actual debate.
  • Talk about the "religious issues" and then try to reach substantive agreement.
• "What is the internet? Is it a duck, or a rabbit?" Two major options
  • "Bellhead": Can answer this question through technical commentary
    • Last mile (connection to consumer) + link to backbone + backbone
    • Defined service levels; network optimized for particular service
    • Lots of intertwinement between telecomm and security/law enforcement
  • "Nethead": End-to-end view
    • The best network is dumb and simply delivers packets
    • Security is a responsibility of the end users
    • "Internet" is a set of standards (e.g. TCP/IP) and relationships (whole is greater than the sum of the parts)
    • Most important: no one is "in charge"
  • VoIP is creating one wedge between these two views
• US situation
  • Telecom regulation is Deeply political, shortsighted, and driven by incumbents (government entities, telecom companies, content companies such as Disney)
  • Network neutrality
    • BrandX - cable deregulated (no obligation to protect neutrality)
    • DSL order - DSL deregulated
    • Clear that a greater choice of providers is needed for competition
  • Telecom arguments
    • Internet broken; need high resources in order to ensure QoS.
    • Expensive infrastructure needs monetization, which needs monitoring, control, and prioritized access.
    • No other network operates neutrally.
    • Target monopolies and deception directly, not through enforced neutrality.
  • Note actual data showing that the US has the least broadband penetration in the developed world.
• Developing nation issues
  • Dealing with incumbent telecom
    • Competitors in Amsterdam and India are laying their own fiber
    • Some are using a microinvestment strategy
    • But note Egypt problem discussed in plenary panel: even access is insufficient, if there's no infrastructure to use and support it.
  • Many argue that most customers only send email, and QoS is needed to make sure that bandwidth is not wasted.
  • The real opportunity is in mobile internet.
• Threats
  • Telecom policy is becoming all of communications policy
  • Choice of viewpoint critical
  • Specific concerns: attach devices, launch applications, and produce and upload contect without permission of company
• Risks
  • Within bellhead: will break the Internet
  • Within nethead: will lose freedom

Caio Pereira:

• Lecturer in Sao Paolo

• Title: Net Neutrality v. Net Deployment in Developing Countries

• General setup
  • Like Canada, Brazil is not really a developing country, but many issues are similar.
  • As discussed by other panelists, the debate in developed world is between Net Neutrality (Nethead) and Net Control (Bellhead).
  • Even though US broadband penetration is low, it's still fairly wide and available. But in the developing world, in many places broadband is a dream, not a reality.
• But the question is still important in the developing world.
  • Deployment is a key issue in the developing world - are the arguments for incentives and recoupment of costs really accurate? It's different when US telcos are arguing for resources for further development.
  • But neutrality may be even more important in the developing world - if there's only one pipe, we need to make sure that it's being used well, and that it doesn't restrict innovation.
• Consequences of the tradeoffs in the developing world
  • A2K framework should say that neutrality is "an important goal"
  • But tradeoffs should not be ignored. In some contexts, they are real, and we must go beyond the "rhetoric" and get to real substantive decisions. Tradeoffs are context-specific, and cannot be treated in the abstract.
  • One clear principle is transparency in bandwidth management (prioritizing).
    • It is especially important in the developing world because of scarcity - if not all requests can be served, then we must know which are lost and why they are chosen.
  • Must discuss discrimination against content and its impact on A2K goals.
    • Some content discriminations are worse than others; it's not an all or nothing problem. Example: discriminating in bandwidth offered to consumers is not the same as discriminating applications.
• Looking at the debate in Brazil
  • Connectivity snapshot
    • Stagnation of telephony
    • Slow growth of wired broadband
    • Explosion of mobile lines
    • Fairly low wired connections of cable and DSL compared to wireless
  • Neutrality?
    • The growing networks, like mobile phones, are incredibly controlled.
      • Different agreements create real private internets
    • TV (DTV + set-top boxes) transitioning to new networks
      • Great concern of development of "walled gardens", private networks
  • Challenges in Brazil
    • Need to bring neutrality debate to changing TV network context
    • Need to bring transparency to all networks
    • Need to separate discrimination in the regulatory discussion
    • Public provision could have an important role through experiments in wireless

Sean O'Siochru:

• Founder of CRIS, Communication Rights in the Information Society

• General commentary
  • Wants to criticize the panel's description - much of it focuses on the developed world, and plans to sort of offload these solutions to the developing world.
  • The principle of network neutrality is important, but it's not yet relevant to the developing world (e.g. Africa).
  • The real concern in the South is for low-cost bandwidth.
• ICTs in Poor Communities
  • There are many problems in poor communities that bar access.
    • Vertically integrated operators restrict access
    • GSM/CDMA are too expensive
    • Regulatory bodies are weak
    • Africa pays twice for data flows
    • Rural bandwidth is mostly from VSAT (wireless), increasing cost tremendously
    • Existing fiber is underused
  • Potential solutions
    • SAT3 in West Africa - underground cable owned by a group of investors who are motivated by short-term gain and protection of vertical integration.
      • It's increasingly proving worse than VSAT, and many are switching back.
    • EASSY consortium in East Africa - also primarily controlled by vertically integrated and/or incumbent operators, and risks running into the same problems as SAT3.
    • "Open Access" campaign - argues for available bandwidth, transparent and affordable prices, and focus on long-term benefit rather than emphasizing return on investment.
      • Also "Public Goods" campaign - make sure that the benefit is for society, not just the developer.
    • Of course, World Bank/EU/ADB could just pay for everything. It would only cost about $200 million. It's simple, and it would work, but it won't happen.
• Moving beyond the cable: National Backbone
  • Possible solutions
    • Open Access - leverage public assets for open access. A good regulator and a good government can replace privatized access and ensure open access. But in practice, getting strong and capable regulators has not been easy.
    • Pro-poor Open Access - deliberately prioritize poor communities. Stop the emphasis on spreading access to those who can afford to pay large amounts for it.
• The Last Mile
  • Wireless IP is "coming of age" - becoming a practical, low cost replacement for switched networks.
  • Creates a great potential for community owned and driven [end-to-end] networks. The whole effort of access is transformed into a development and empowerment movement, in which the community itself structures its access.
• For reference: http://www.propoor-ict.net

Questions:

Q1: What about the possibility of arbitrage/tunneling to avoid content discrimination? How is the political economy of this issue different in the developing world, which has fewer resources for enforcement?

A (Susan Crawford): There are a variety of responses to this. The answer in the developed world is one of giving up - people will always route around. A (O'Siochru): This raises the question of private v. public owned networks. The answer is to avoid private networks altogether. A (Pereira): Always take into account the specific circumstances. It's a pervasive problem in all contexts.

Q2: A story of the expense of access for an NGO, noting other issues not raised by the panel concerning actual obstructions to access.

A (O'Siochru): Has published a report with examples detailing these barriers and how a government can make access happen (though it's very difficult).

Q3: Public good can be separated from public ownership.

A (Pereira): Government can get into these issues in many different ways. A (Geist): Canada has money to expand broadband access, and is planning to essentially supplement private companies' revenue from poorer communities.

Q4: Can the FM Radio experience in Africa tell us anything about access?

A (O'Siochru): The areas are very different, and it's hard to borrow solutions from the radio movement. Radio and internet are being seen as part of the same package, though. There are parallels, but no real lessons per se.

Q5: Sometimes it is impossible to get the state to act. The mobile phone companies are quite active; the largest company in Ghana is a mobile phone company.

A (O'Siochru): The private companies will be an important component of the final solution.

Q6: How do you undo the legacy of mobile phone companies?

A (Geist + Godwin): We are approaching a private monopoly situation. We haven't really worked the numbers to determine when the companies have recovered their investment and when they possess a monopoly.

Powerpoints

• Carlos Pereira: PPT : PDF
• Michael Geist: PPT : PDF
• Susan Crawford: PPT : PDF
• Sean O'Siochru: PPT : PDF

Papers and resources