Interconnection is Different for Internet than Railroads or Electricity – Part 55 FCC Open Internet Order Series

Published June 27, 2014

Some things are way too important to let slip by uncontested.

The FCC has asserted a foundational regulatory premise that warrants rebuttal and disproving, given that the FCC is considering if Internet access, and Internet backbone peering, should be regulated like a utility under Title II telephone common carrier regulation.

In an important speech on Internet interconnection last month to the Progressive Policy Institute, the very able and experienced Ruth Milkman, Chairman Tom Wheeler’s Chief of Staff, asserted that “communications networks are no different” than railroad and electricity networks when it comes to interconnection. “… At bottom… the fact is that a network without connections and interconnections is one that simply doesn’t work. Disconnected networks do not serve the public interest.”

The grand asserted regulatory premise here is that because communications networks are “no different” than railroad or electricity networks, they should have proscriptive government regulation to ensure that they are, and remain, interconnected to ensure that the public is protected.

If this sweeping assertion is accepted at face value without challenge, the FCC could have unfettered incentive and justification to begin regulating Internet backbone peering for the first time.

The facts are that this FCC foundational assertion about communications interconnection being no different than railroad/electricity interconnection — is fundamentally untrue.

Internet communications networks are completely different than railroad and electricity networks and the Internet backbone has worked successfully and almost flawlessly for two decades without FCC regulation.

How are Internet networks completely different than railroad and electricity networks?  

First, railroad and electricity interconnection is place-dependent, Internet “interconnection” is not place or physical-location-dependent.

This is a huge difference as physical-place-dependency can create a physical interconnection chokepoint in railroads or electricity. In contrast there are no physical-place-dependent chokepoints for the Internet because one can access/connect to the Internet from many different places, through many different entities, and via many different technologies, e.g. electrically via wires like copper or coax, optically via different fiber configurations, or wirelessly via many different licensed and unlicensed frequencies and providers.

Choice of place, facility, provider, and technology mean multiple dimensions of competition and no lasting chokepoints because if one encounters a temporary congestion problem in one part of the Internet, one has the choice to take their traffic and business elsewhere. No chokepoints mean no need for proscriptive regulation of Internet peering arrangements.

In the Netflix example, Netflix has a wide variety of choices (by place, facility, provider, or technology) to connect to any other Internet network, whether it be one of many CDNs or transit providers, or directly with a network provider. Netflix’ complaint is not over a chokepoint interconnection problem, but that it does not want to pay anything to ISPs for the costs of sending 34% of the Internet’s downstream traffic.

Netflix maintains, under its self-serving re-imagination of “net neutrality” that the FCC mustpermanently mandate a price of zero for Netflix traffic so users are forced to shoulder the entire cost burden of Netflix’ 34% of downstream Internet traffic.

Second, railroad and electricity interconnection is hardware-dependent, whereas Internet interconnection is software-dependent.  Railroads and electric networks require one universal physical standard of wheel gauge and axle width, or physical electrical transformers and wall plugs, to interconnect to these respective networks. In contrast, the software design and protocol of Internet connections make interconnection hardware-agnostic, seamless and automatic, and hence inherently competitive and choice-rich.

Simply, the genius of Internet packet-technology networks is that they do not require any interconnection, permission, or negotiation points, because inherent in Internet Protocol is that packets are automatically routed seamlessly between different internet networks to their destination by design. Inherently Internet packet technology makes the concept of telephone interconnection obsolete because the technology supplants what used to require hardware and regulation to achieve. Most simply, Internet protocol innovation inherently obviates an FCC role for regulating Internet backbone regulation.

Third, railroad and electricity interconnection involved analog technology, whereas Internet interconnection involves digital computer technology. Importantly, railroads required a setcontinuous physical path or circuit from point A to point B. Electricity networks require acontinuous electrical circuit from origin to destination.

In contrast, digital technology in general, and Internet packet technology in particular, isdiscontinuous — the antithesis of a telephone or electrical continuous circuit. It is this inherently discontinuous digital innovation that enables Internet networks to be place-agnostic and hardware-agnostic, and hence inherently competitive and choice-rich.

More specifically, the innovation of digital IP packet networks subdivides information into many small packets to enable more efficient transmission. The packets get individually routed unpredictably and comingled with other packets to minimize bandwidth waste. At the ultimate destination, the packets get immediately reassembled by any device anywhere. Internet Protocol is inherently a competitive technology, made even more competitive inherently by Moore’s law, which ensures that digital networks continually enjoy rapidly declining computing costs.

In sum, the Clinton Administration knew when it privatized the Internet backbone twenty years ago that it did not require FCC involvement, and that it should not be subject to Title II common carrier regulation of prices, terms and conditions.

Twenty years of phenomenal success — where the competitive Internet backbone continually adapted exceptionally to handle the exponential growth of Internet traffic without material or lasting incident — is overwhelming evidence that the place-agnostic, software-driven, digital Internet backbone does not need any type of utility interconnection regulation.

Don’t let anyone assert unchallenged that Internet interconnection is no different than railroad or electricity interconnection. If that patently untrue assertion – that interconnection will not happen without government — is unchallenged, it enables regulators to justify unnecessary, unwarranted, and unjustified regulation of the Internet backbone.

The FCC does not need to regulate or intercede in Internet peering disputes, because if a company like Netflix does not like the prices, terms or conditions, offered by an ISP, they have the competitive choice to negotiate with any number of CDNs or transit providers to deliver their traffic to users.

So since there are so many CDN and transit choices, by definition a peering dispute at a particular place on the Internet cannot result in the “disconnected network” problem FCC Chief of Staff Milkman apparently feared in her recent speech on the subject.

The success and growth of the unregulated model for the Internet backbone peering marketplace has been nothing short of phenomenal in enabling and ensuring everyone reasonable connection to the Internet.

The Internet backbone peering marketplace works near perfectly. As the old adage says; “if it ain’t broke don’t fix it.”


FCC Open Internet Order Series

Part 1: The Many Vulnerabilities of an Open Internet [9-24-09]

Part 2: Why FCC proposed net neutrality regs unconstitutional, NPR Online Op-ed [9-24-09]

Part 3: Takeaways from FCC’s Proposed Open Internet Regs [10-22-09]

Part 4: How FCC Regulation Would Change the Internet [10-30-09]

Part 5: Is FCC Declaring ‘Open Season’ on Internet Freedom? [11-17-09]

Part 6: Critical Gaps in FCC’s Proposed Open Internet Regulations [11-30-09]

Part 7: Takeaways from the FCC’s Open Internet Further Inquiry [9-2-10]

Part 8: An FCC “Data-Driven” Double Standard? [10-27-10]

Part 9: Election Takeaways for the FCC [11-3-10]

Part 10: Irony of Little Openness in FCC Open Internet Reg-making [11-19-10]

Part 11: FCC Regulating Internet to Prevent Companies from Regulating Internet [11-22-10]

Part 12: Where is the FCC’s Legitimacy? [11-22-10]

Part 13: Will FCC Preserve or Change the Internet? [12-17-10]

Part 14: FCC Internet Price Regulation & Micro-management? [12-20-10]

Part 15: FCC Open Internet Decision Take-aways [12-21-10]

Part 16: FCC Defines Broadband Service as “BIAS”-ed [12-22-10]

Part 17: Why FCC’s Net Regs Need Administration/Congressional Regulatory Review [1-3-11]

Part 18: Welcome to the FCC-Centric Internet [1-25-11]

Part 19: FCC’s Net Regs in Conflict with President’s Pledges [1-26-11]

Part 20: Will FCC Respect President’s Call for “Least Burdensome” Regulation? [2-3-11]

Part 21: FCC’s In Search of Relevance in 706 Report [5-23-11]

Part 22: The FCC’s public wireless network blocks lawful Internet traffic [6-13-11]

Part 23: Why FCC Net Neutrality Regs Are So Vulnerable [9-8-11]

Part 24: Why Verizon Wins Appeal of FCC’s Net Regs [9-30-11]

Part 25: Supreme Court likely to leash FCC to the law [10-10-12]

Part 26: What Court Data Roaming Decision Means for FCC Open Internet Order [12-4-12]

Part 27: Oops! Crawford’s Model Broadband Nation, Korea, Opposes Net Neutrality [2-26-13]

Part 28: Little Impact on FCC Open Internet Order from SCOTUS Chevron Decision [5-21-13]

Part 29: More Legal Trouble for FCC’s Open Internet Order & Net Neutrality [6-2-13]

Part 30: U.S. Competition Beats EU Regulation in Broadband Race [6-21-13]

Part 31: Defending Google Fiber’s Reasonable Network Management [7-30-13]

Part 32: Capricious Net Neutrality Charges [8-7-13]

Part 33: Why FCC won’t pass Appeals Court’s oral exam [9-2-13]

Part 34: 5 BIG Implications from Court Signals on Net Neutrality – A Special Report [9-13-13]

Part 35: Dial-up Rules for the Broadband Age? My Daily Caller Op-ed Rebutting Marvin Ammori’s [11-6-13]

Part 36: Nattering Net Neutrality Nonsense Over AT&T’s Sponsored Data Offering [1-6-14]

Part 37: Is Net Neutrality Trying to Mutate into an Economic Entitlement? [1-12-14]

Part 38: Why Professor Crawford Has Title II Reclassification All Wrong [1-16-14]

Part 39: Title II Reclassification Would Violate President’s Executive Order [1-22-14]

Part 40: The Narrowing Net Neutrality Dispute [2-24-14]

Part 41: FCC’s Open Internet Order Do-over – Key Going Forward Takeaways [3-5-14]

Part 42: Net Neutrality is about Consumer Benefit not Corporate Welfare for Netflix [3-21-14]

Part 43: The Multi-speed Internet is Getting More Faster Speeds [4-28-14]

Part 44: Reality Check on the Electoral Politics of Net Neutrality [5-2-14]

Part 45: The “Aristechracy” Demands Consumers Subsidize Their Net Neutrality Free Lunch [5-8-14]

Part 46: Read AT&T’s Filing that Totally Debunks Title II Reclassification [5-9-14]

Part 47: Statement on FCC Open Internet NPRM [5-15-14]

Part 48: Net Neutrality Rhetoric: “Believe it or not!” [5-16-14]

Part 49: Top Ten Reasons Broadband Internet is not a Public Utility [5-20-14]

Part 50: Top Ten Reasons to Oppose Broadband Utility Regulation [5-28-14]

Part 51: Google’s Title II Broadband Utility Regulation Risks [6-3-14]

Part 52:  Exposing Netflix’ Biggest Net Neutrality Deceptions [6-5-14]

Part 53: Silicon Valley Naïve on Broadband Regulation (3 min video) [6-15-14]

Part 54: FCC’s Netflix Internet Peering Inquiry – Top Ten Questions [6-17-14]

[Originally published at Precursor Blog]