A Practical Navigator for the Internet Economy

Reframing Policy
by Reclassifying IP as Telecommunications

Download entire issue

Summary of pp. 1-29

Erik Cecil sees network neutrality as an imaginary solution to the problems of telecom deregulation.  Lee Selwyn in a brilliant June 2009 presentation showed the lunacy of the FCC’s deregulatory infatuation. “FCC deregulated for deregulation’s sake. Confused MEANS with ENDS. No goal other than deregulation itself.  No benchmark for judging success.  No process for ex post evaluation of outcomes.”

Selwyn very clearly establishes how, although the extent of the natural communications provider monopoly during the period of time from Carterphone to Powell Martin shrunk, the fact that there was essentially only one point of entry for customers into increasingly closed networks created a meaningless choice between the telco or cable co. Since the FCC also moved all internet protocol services under the Enhanced Service Provider emption which was unregulated, the incumbents were given the ability to recreate the environment of pre 1984 monopoly services where the best the customer could hope for was a tweedle dum versus tweedle dee choice between cable co and telco – both with similar services and prices.

Erik, first in his blog and then in more detail in a COOK Report interview argues that that network neutrality will leave this fundamental regulatory injustice untouched. Eric finds that Network neutrality is well-intentioned manipulation of existing FCC rules. NN tries to create common carrier rights out of an exception to common carriage.  IP was regulated as an exemption to common carriage as a special case to protect it even before the 1984 breakup.  The ESP exemption was originally done as a means of forcing ATT the wire owner to interconnect with IBM the device owner and to interoperate on more or less equal terms without cannibalizing IBM

Decade after decade the interest of the network owner (the wire) has been to extract as much payment. as it possible can from the device owner. As devices have gone digital and increased by many orders of magnitude in their  power and the networks have taken a short term rent extraction point of view, the device owners justly complain that the wire owners are using a system codified in 1934 to cripple their ability to contribute to the general well being of the economy.

Regulation has been an attempt to balance the competing interests of device owners against wire owners.  When the devices owners at the edge create new values – the internet for example - the wire owners always seem to be able to play regulatory jujitsu to extract exorbitant value from the creativity of those who invest in improving devices while the wire owners generally invest in ways that ensure that only they extract value from the creativity of the device owner.

The original IP – ESP exemption was a policy choice to encourage competition.

But everything that counts now is IP traffic – and IP traffic is exempt from regulation. And since they own the access points for IP traffic into and out of their networks, they have flipped regulatory ESP exemption on its head so that its continuation ensures the instantiation of the monopoly of the facilities owner in perpetuity.

Until you find a way to stop the network owner from extracting monopoly rent from its wires, you will never cross the transition chasm into the widespread productive use that Carlotta Perez speaks of as the final deployment phase of the ICT revolution.  Until these technologies go into widespread use, you don’t get all their economic benefits. Under wireline owner-extortion you can kiss device innovation good bye.

The wire owner is a predator.  Now why would you put a private entity in control of building, constructing, and operating the wires?  Your first answer is going to be its ok to do because they are regulated.

But regulators are ordinary people who are subject to the economic lobbying pressure of the wire owners.

But nothing changes because we have public officials elected with private money in charge of making sure that private ownership of public property (the right of way and the infrastructure we’ve funded for a century) is used in the public interest.   But Play that out over time; the private interests, by necessity, are always stronger than their public overseers.

What we can help the FCC do is articulate a vision that says – why don’t we just make all this into big utility infrastructure and, if we can do that, then what we can do is get some regulatory control and leverage back to people who really need it.   We want to get the states out of their hair because the Bells totally own the state Commissions.

Change needs to be anchored in statutory bedrock rather than rulemaking to get our strategic direction back to a sound direction on a national level. In order to scale and in order for cloud computing to succeed, devices need to interact uniformly with the wires no matter what state they happen to be in. You need to open up all of this uniformly by calling all of it telecommunications in your statutes. Do that and then all of it is common carriage and if it is common carriage, because it is also telecommunications, the FCC has jurisdiction to preempt state regulation because its inter state character is recognized.

The solution is for the FCC to call those IP or “net neutrality” bits telecom and thereby exempt them from state regulation.  Do that as a matter of statute and, when you are done, we can have a rule making,  and in the rules you can say a network operating under common carriage must kill Spamford because any carrier is allowed to exercise reasonable control over its network.

If end-to-end is inter state and if the FCC says it is common carriage, then anyone with a device that connects to the network gets the rights of common carriage.  With my device attached to the net, I don’t need the carrier to sell me any service other than a big symmetrical pipe.

How Do You go about Getting IP removed from ESP and treated as Telecommunications?

Marc Cooper affirmed that this could be done and on page 15 Chris Savage listed a nine point argument that would permit it to be done

Since IP could be re-classified as telecommunications. That would help because it would make it far easier to end the regulatory gaming process by the wire owners. How to start such a ball rolling?

First:  Sue the incumbents

Second;  Sue the FCC. Net Neutrality will be appealed. If someone does not set this up in Comments to the FCC in the relevant proceedings there is less of a chance of meaningful change.  There will be a shootout at the DC Circuit.  Those who are positioned in advance will win. Consumers, however, will most likely lose.

Third: Hold the administration's feet to the fire. Do not believe anything anyone says until you see it implemented. The road to 2009 was paved with good intentions and things said behind closed doors. If they cannot say it and embrace it in the open, it is not real.  

The problem is since the regulatory process has been successfully gamed in the interest of the incumbents rather than the American people and we have an increasingly expensive and antiquated telecommunications infrastructure -

We might ask  who will be the new Theodore Vail getting a legal monopoly and antitrust immunity for rate of return regulation of a gold-plated infrastructure?  Promises of a gigabit to every home...

Why not Google?

Here’s their deal for the customers: free telephony, free broadband, non discriminatory access to the world's knowledge in return for rate of return regulation......(And the freedom to make money out of ad streams as well).

All that is needed is to submit to being a regulated utility, allow the consequences to undermine the value of the competition, buy the underlying communications infrastructure for a song (but safeguard the pension rights to ensure political credibility). . .manage congestion and secure free speech...  Where's the downside?

Regarding dark fiber networks, the real point there is not so much that Google has one as it is that there is NO LOOP and NO BACKBONE. There is just network.  Where consumers and individuals are getting screwed badly is that we continue to fund loop providers NOT to DEPLOY advanced equipment - wide open fiber optic.

The asymmetrical regulatory system that they control won’t allow this to change because the basic business models for the incumbents are still predicated on economically milking their control over access to the network by all of us since this is the one place where they still enjoy a monopoly.  Their rates, taxes, and subsidies are all based on the costs of the copper based “local loop,” over which, as Selwyn has shown, is the only thing that they still have complete control.

While the heart of the network is predicated on energy efficient and transport unlimited fiber, they have used regulation to take over and turn the system inside out that at one point was used to protect the interests of their customers.  While they may want to be in Google’s business, they are holding on for dear life through a regulatory system fractured amongst 50 state PUCs to a business model predicated on scarcity and charging for every unit of “service.”

As Erik Cecil concludes the problem, ultimately is not lack of regulation but the asymmetrical ESP regulatory exemption that was originally done to keep the incumbents from killing internet protocol based packet switching.  However, with the incumbents using the alleged competitive nature of IP services as a means to gain removal of their common carriage obligations from the last vestige of their remaining monopoly – the local loop - because internet as an enhanced service is still alleged to be competitive.  The only problem is that they now control every entryway into that unregulated “enhanced services” space and so are free to do what ever they like to anyone wishing to compete.

Symmetrical Megabit per Second Wireless  to Mt Everest, pp. 30-45

We interview Pavan Shakya who in partnership with Pemba Sherpa of Namche Bazar and WordLink has a Canopy radio based wireless link from WorldLink in Kathmandu in three hops to a ridge top known as Rautah at 6,500 feet and from there 98 kilometers to the Everest View Hotel at Syangboche airstrip at 13,500 feet overlooking Namche Bazaar.  Benefitting from a connection to fiber made by the Nepal electric authority at the Indian border and running to Kathmandu they are bringing symmetrical one megabit per second bandwidth for $700 a month as opposed to $1300 for 384 by 256 kbs Vsat bandwidth.

Using a total of 20 photographs and maps we explain the topology of the network and examine its uses which range from communications support for sherpas and trekkers to climate change research.

IP v4 Exhaustion what is at stake – pp. 46 -63

With 90% of Ipv4 block now assigned by the routing registries we had another long discussion.  of possible work arounds as well as of the issue of what markets for reassignment of IPv4 addresses would look like.

Variable outcomes seem to be how many new users  can be accommodated by increased use of Network address translation and whether the development of a market for IPv4 addresses would be able to use monetary incentives  to free up enough unused IPv4 addresses to supply the needs of new entities wanting to supply internet service.  It was pointed out that such addresses would need to be routed to be worth anything.

The existence of IPv4 addressing as Tom Vest points out serves as a unit of currency that enables a new business to get and use the ASN number necessary to use the internet as an eco-system.  No matter what technology a new entrant has and no matter how attractive its business plan, without an ASZN number and routable IPv4 bloc or blocs and new business may not participate in the internet ecosystem. 

Tom puts the problem this way: During the period between (a) the exhaustion of the unallocated IPv4 pool and (b) the day when possession of IPv4 ceases to be an absolute prerequisite for enjoying "autonomous" status as a provider of routing or other Internet-based services, the prospect "competitive entry" will cease to exist in all countries -- or at least, in all countries that are enmeshed in what we call "the Internet" today. New entrants may continue to emerge during that period, but each new entry will be contingent on securing the permission of incumbent services providers (and also, implicitly, of competing speculators or "market makers"), in the form of an " IPv4 transfer" of some kind. My guess is that "fee simple" IPv4 transfers -- the only kind which would provide the same degree of commercial freedom of action that incumbent IPv4 recipients have always enjoyed -- will be both very expensive and rare.

Later Chris Savage asked and Tom Vest replied:

Savage: I'm not making any predictions here about how "the market" for addresses would work out, but the fact that different businesses with different applications will obtain differential benefits from IPv4 addresses, being behind a NAT, and IPv6 addresses suggests that some kind of trade/market mechanism could actually improve things, in the absence of some Leviathan directing everyone to move to IPv6 or imposing binding rules on the sharing/reallocation of remaining IPv4 addresses.
Am I missing something?

Vest: If you are, it's only a full appreciation of how amazing and improbable is the thing that we're about to give up. This is only the second time in human history that a completely novel liquidity mechanism has emerged out of the spontaneous interaction of independent economic agents. It represents a model of how liquidity systems can and *should* work under conditions of "abundance," e.g., in a world dominated by nonrival real factors. The fact that it has spawned such a vast, diverse, and thriving/innovative economy in the *absence* of a pervasive/continuous monetary payment mechanism represents a direct challenge to the world views of many of the more reductionist schools of economic thought -- and this is one of the reasons why this particular view induces such an angry reaction in some quarters. Consciously or unconsciously, the reductionists recognize it for what it is: a description of a world in which their entire understanding of what motivates people, and how economies work, has been falsified (or at least de-universalized).

And later Tom Vest concludes

I'm not fundamentally FOR anything in particular, except the continued survival, growth, and fruitful evolution of Internet. I'm not fundamentally AGAINST anything in particular, except things that would make the above very unlikely or impossible.

My own work is not explicitly normative, and to the extent that is has normative implications, they go far beyond the issue of addressing format migration. I've posited what I claim is a strong, empirically sustained theory about what the Internet "is" at the most foundational level -- e.g., how it came to possess that key feature/functionality that makes it unique and important, how that particular feature imposes certain unavoidable structural vulnerabilities, etc., etc. What I further claim is that the isomorphism between the Internet and another implementation of the exact same functionality -- the latter of which has been carefully observed and documented through many centuries of growth, evolution, and not-infrequent collapses and recoveries -- makes my own work (very) unusually useful for predictive purposes. Lots of partial, transient "mimetic" isomorphisms have been identified by economists and policy analysts in various fields -- mostly artifacts of the application of "lessons learned" in one domain to similar problems in another -- but as far as I can tell, this is the first solid example of purely organic, naturally occurring, and durable isomorphism that has ever been identified.  For what Tom is talking about see the detailed report on his exposition of IPv4 addresses units that are isomorphic to units of currency within the banking system of all modern economic systems in the December 2008 Cook Report.

So it's predictive, and having a very personal interest in seeing the Internet survive, I can't help but make as strong a case as I can AGAINST a course of action that, history suggests, will lead to (at least one or several possible) non-survival outcomes. This is not merely (or self-evidently) a cry in vain, however. The Internet community was able to self-organize itself enough to implement an ingenious hybrid solution to the last ecological catastrophe that it faced (c. 1990-1993). Similarly, of the various non-sovereign, self-governing "banker's clubs" that defined and managed the liquidity/monetary policy mechanisms of their (late 18th-19th c.) day, many confronted and were obliged to overcome various systemic challenges that could have caused them to collapse. Although none have survived to the present day, many managed to adapt successfully and to thrive for many decades.

I wouldn't wish to see any of the four* critical features of the Internet weakened or eliminated, and so I've chosen to focus 100% of my own personal efforts on the only strategy that IMO could achieve that goal, and so obviate any need to play triage and, ultimately, sacrifice one or more in an attempt to save the others. That strategy involves convincing the self-governing parties themselves to do (at least some of) the right things, and to not do (even one of) the fatally wrong things.

So far, I would not say that I have been entirely successful; regardless, I know of no better alternative than to keep trying.

I'll close with a corrected* version of the formulation that I gave in a response to Fred last week:

Of the four* distinct (and unprecedented, and immeasurably valuable/important) conditions that are sustained by the current hybrid system of Internet technologies and institutions -- those being (1) competitive, open, not-"fixed" markets, (2) industry self-governance, (3) transparency to most formal and informal "international" trade, investment, and communications-related barriers, and (4) it actually works -- _choose one_, or maybe two. Or if you're really, really optimistic, choose three and cross your fingers. Better make your choices soon, however, because a world without IPv6 -- or some other immediate and equally open, enabling, and functional method of Internet attachment -- is very unlikely to allow all four to exist at the same time.

So you see, given that particular mix of goals, I'm really just doing the only thing that I can do -- that I must do.



Reframing Telecommunications Policy
Erik Cecil Describes Why Net Neutrality Doesn’t
Solve the Real Problems and Offers an Agenda
that Is Implementable and Would Free the Economic
Productivity of Device Owners

Introduction and Summary of Erik
Cecil’s Position                                    p. 1
The Origin of the Problem                            p. 2
What Net Neutrality Is                                    p. 7
Device Owner Versus Wire Owner                            p. 7
Device Versus Wire Encore                                p. 8
Where Regulators Get Their Power                        p. 10
Common Carriage                                        p. 12
The Next Chapter                                        p. 14
Smashing the Conventional Paradigm                    p. 16
Who Will Be the New Theodore Vail?  -- Google?            p. 24

Canopy Wireless - Kathmandu to Everest
Entrepreneurial Skill on Verge of Bringing Affordable Internet to Sherpas as well as Trekkers

Editor’s Introduction                                        p. 30
Solu Khumbu Business Model                            p. 34
At Syangboche                                            p. 40

Symposium Discussion  September 16 - October 22

What is at Stake as IPv4 Runs Out

Delayed Binding?  Use of Names Rather Than Numbers?    p.50
In Search of the Scalable                                    p. 52
What Markets Might Look Like                            p. 52
Getting Routed                                            p. 54
How TCP/IP Decoupled the Telecom Bottleneck            p. 59

Executive Summary                              p. 64