My Insanely Long Field Guide To Understanding FCC Chairman Tom Wheeler Statement On Peering.

At the press conference following the Federal Communication Commission (FCC) March 31 Open Meeting, Chairman Tom Wheeler made the following observation:


“Interconnection is part of the Network Compact.” Peering “is just a $3.50 word for interconnection.”


Wheeler followed up this statement by explaining that there was a difference between “network neutrality” and the “open internet” on one hand and “interconnection” as the ‘path to the Internet’ on the other hand. While government has a critical role in monitoring peering/interconnection to protect the values of the Network Compact, it isn’t a network neutrality issue. You can see Wheeler’s full statement here (Start at 144:45 – 147:23 has unrelated stuff in middle) (transcript here).


After the meeting, the FCC released a separate statement that they really mean it when they say that they aren’t going to do peering as part of the Net Neutrality rules. While Brendan Sasso at National Journal gets points for noticing that ”the FCC could decide to enact separate regulations on the issue or force Comcast to accept new rules in order to receive permission to buy Time Warner Cable,” most folks I’ve read in the press have broadly interpreted this as indicating the FCC will not look into the Comcast/Netflix dispute or complaints by Cogent and Level 3 about large edge-providers squeezing them for higher interconnection fees.


Personally, I think most people are totally misreading this. Wheeler’s statements make it look more likely to me that the FCC will start looking closely at the Internet peering market, not less likely, especially as part of the Comcast/TWC deal. Indeed, Comcast’s Chief Lobbyist David Cohen, who ranks in my book as one of the absolutely smartest and most effective telecom lobbyists ever, has already started backing away from earlier statements that regulators would ignore peering issues and that he expects them to look at the Comcast/Netflix deal. (Unsurprisingly, Cohen also said he expects regulators to find no problems with the deal and called Netflix CEO Reed Hasting’s arguments that this eviscerated net neutrality “hogwash.”)


Below, I will rant at considerable length that (a) Wheeler is right, this is not a “network neutrality” issue, but the same goddam interconnection issue that we have struggled with for more than a hundred years in every networked industry from railroads to electricity to broadband; (b) The FCC needs to actually look at this and study it and understand how the market works before it makes any decisions on what to do; and, (c) While Wheeler is not saying in any way, shape or form he actually plans to do anything before he has real information on which to base a decision, he is signaling — for anyone actually paying attention — that he is, in fact, going to actually look at this as part of his overall transition of the agency around his “Fourth Network Revolution” and “Network Compact” ideas.



While this last would seem pretty basic and obvious, it represents a significant change in policy from the previous insistence that IP magic pixie dust obscures all things Internet and makes them invisible to the FCC. Whether I agree with what Wheeler ultimately does or not — and I have no idea what he might ultimately do here, he could decide the market is competitive and working just fine — I don’t believe Wheeler is going to go around with his eyes and ears covered blathering about the magic nature of the Internet. I think Wheeler is actually going to check under the hood and see what actually makes the damn thing tick — and Comcast is just the company to help him do it.


Much ranting below . . .



Since the Comcast/Level 3 blow up in late 2010, I have pretty consistently complained that “peering” simply represents the same sort of problem we’ve always had in the telecom world of whether last mile providers (or some other point in the distribution chain) end up with disproportionate power because of network effects, termination monopoly issues, or other factors that make network economics work differently from general markets. As I wrote in this blog post in February 2011:



“Because from a telecommunications perspective, it looks all the world like Comcast demanded a termination fee from Level 3 and got it. Yeah, yeah, Level 3 pretending to be CDN blah blah blah. It still looks like a demand for an unregulated termination fee, and you can bet that every single residential ISP looking for new revenue streams (and aren’t all ISPs looking for new revenue streams) has taken notice and is watching to see whether Comcast can make it stick. For their part, backbone providers have to be thinking about how they can recoup their new expenses from smaller ISPs who depend on them for traffic. Those of us that remember the old days of CMRS or CLEC negotiation for termination agreements may lick their chops or shudder, depending on whether or not they get paid by the billable hour, at the prospect of a free for all emerging in this space.


“Maybe it will all work out. Maybe it was just Comcast or L3 being greedy rather than a tremor from the grinding together of forces beneath the surface. But I’m betting on structural issues and greed to mean more tremors and increasing demands from a variety of stakeholders for the FCC to at least start looking at the issue.”



Wheeler reads the lessons of history the same way, and raises these concerns every time he talks about the history of network revolutions. His E-Book “Net Effects: The Past, Present and Future History of Our Networks — History, Challenges and Opportunities,” lays out what he thinks are the general lessons from the history of previous “network revolutions,” how they apply to the evolution of the Internet, and the appropriate role for the FCC. In particular, he has an entire section on interconnection and interoperability that I urge folks to actually read if they want to understand Wheeler’s overall philosophy and approach. In particular, Wheeler makes a critical difference between regulating “the Internet” and the role of the FCC in ensuring things like interconnection and universal service happen:


“We must be clear. “Regulating the Internet” is a non-starter. What the Internet doesis an activity in which policy makers should not be involved (other than assuring overriding purposes such as the ability to complete 911 calls or the ban of child pornography). Regulating Internet access is a different matter. Assuring the Internet exists as a collection of open, interconnected facilities is an appropriate activity for the People’s representatives.”


Which I get is a distinction most people don’t understand, and that broadband access providers deliberately like to obfuscate. It’s much easier if we draw the simplified picture of a single user, a dotted line leading to a magical cloud, and then another dotted line leading to another edge user or provider. But there is no magic cloud. It is, to invoke the late Senator Stevens, “a series of tubes.” How those tubes connect to one another is a different question from what sort of crap flows through them in the first place.

I’ll add Wheeler is hardly alone in drawing the same lessons from history. I’ve been going on about this for almost 15 years. Tim Wu draws similar lessons in “The Master Switch” when he talks about the repeated pattern he calls “The Cycle. Mark Cooper has written a bunch of stuff about this over the last decade or so, as have others. While not a universally shared view, it’s not whacko cuckoo bird stuff either.

In direct contrast, we have the folks who argue that because the Internet is new we ought to utterly reject the idea that anything we learned before ought to get junked and thrown away. We hear this constantly by those disparaging the Communications Act (and the FCC generally) as old, and therefore useless and utterly unsuited to the current networked world because obviously anything so thoroughly grounded in the past cannot have any relevance to the future.

I will characterize this as the debate between the George Santyana (“those who cannot remember the past are doomed to repeat it”) school v. the Henry Ford (“history is bunk”) school . Or, because I can never resist a good quote from Ecclesiastes: “If there is a thing of which they say ‘Behold it is new!’ It has been already in the ages that were before us.” Ecl. 1:10.


Fact Based Policy Ought To Be Based on, Well, Facts.

Knowing that bad things could happen doesn’t mean we ought to jump right to regulation, or that we can automatically know the right answer. At the same time, neither I or my employer, Public Knowledge, have actually suggested that the FCC should jump directly to some kind of regulation. Instead, as captured in this 2010 video on the Comcast/L3 spat and this detailed economic analysis we submitted to the FCC in 2011 in an utterly vain effort to push them to get their head out of their ass, what we want is for the FCC to friggin’ do its job and actually figure out how this market WORKS.  Not how lots of conflicting people tell you on websites how it works. Not a vague 50,000 overview or even a more detailed summary like this one here.

Because, and I will know this will come as a shock to people, in a complex and opaque market like this one, no single individual or company can know the whole story. Really. I don’t care who you are and how long you’ve worked in this business or whatever. All any individual has, and I am all too painfully aware that I am talking about myself here as well as anyone else, is a slice of the puzzle with a highly simplified view as to how the other slices fight together — with probably a few pieces missing.

And understanding how this stuff works is really important. Because, as we have seen with rural call completion, as we saw on Fire Island when Verizon decided they could roll ahead with the beta version of Voice Link to replace their copper network, these networks can be both incredibly resilient and incredibly fragile at the same time. There are all sorts of working parts that lie under the surface and buried things people installed and forgot about that — if you actually know how the system works — you can manage. But if things start to go wrong and you don’t have a clue then you don’t even notice things are going wrong until they go very, very very wrong.

And yes, even without calling this a Title II telecommunications service, the FCC has both the legal authority to find out how the market works and — I would argue — the legal responsibility to find out how it works. (What it can do once it figures that out if it finds problems or potential problems, is a whole ‘nother question.) Kindly take a look at 47 U.S.C. 154(i), 154(o), 162(1)218257303(g)403521(4)533(c)13011303 and, of course, everyone’s favorite Section 706 of the 1996 Telecommunications Act, codified at 47 U.S.C. 1302.


So Why Doesn’t Wheeler Just Start Investigating?

I will save my general rant about the War for Willfull Ignorance for another time. Suffice it to say for now that industry and the other worshippers of the Gods of the Marketplace have worked very hard to make it “highly controversial” to even think about the FCC actually looking at anything, worked hard to eviscerate the tools for collecting information, made the act of investigating how anything Internet related tantamount to regulating the Internet with all manner of blather about ‘burdensome costs,’ ‘regulation by raised eyebrow,’ agencies trying to preserve their power blah blah blah. At the same time, all regulation must be ‘fact based.’ And, of course, by ‘fact based’ we mean precisely the sort of broad industry wide information the worshippers of the Gods of the Marketplace have made it heresy and an abomination to collect.

Anyone who thinks I am exagertaing should look at the recent manufactured dust up over the FCC’s survey of how communities get news.  The FCC tried to gather up exactly the kind of information opponents of media regulation say we need to have — evidence about precisely how communities get exposure to news and diverse views. Yet through the magic of the conservative noise machine, this little voluntary survey (yes, it was voluntary) got transformed in the mass media (coincidentally controlled by the same people who want ownership limits abolished) into a full on assault on the First Amendment.

Wheeler is not stupid. He knows that if he just up and said “Sure, we’ll launch an investigation into peering and stuff” he would get the same treatment.  Also, you don’t just whip this stuff out overnight. It takes staff awhile to actually write something useful, figure out what questions to cover, and even then — in a standard notice of proposed rulemaking — you only get what companies chose to voluntarily share. So anyone who files evidence of shenanigans and bad conduct are dismissed as whiny competitors seeking regulatory advantage blah blah, while the evidence submitted by the incumbents is dismissed as a whitewash hiding all the good stuff, blah blah. and staff get caught in the middle getting yelled at no matter what they do.

At the same time, Wheeler can’t just drop this stuff out of the blue. Not only does he need to encourage folks in industry and the public who care about this to come forward with relevant information, having the FCC suddenly change course and announce they will investigate prompts its own sort of pushback. “What! The FCC has never even looked at this before! How can you even think you have authority to look at it?” Everybody in the industry freaks out, Wall St. analysts start going all weird/crazy, frogs fall out of the sky, etc. The Republicans scream at how they are taken by surprise and cut out of the process. Finger wagging from Congressional Republicans! Regulatory overreach! etc.


The Ben Bernanke School Of Signaling.

Observing Wheeler’s approximately six months as Chair so far, we find that he basically employs the same idea about “regulatory signalling” as former Fed Chair Ben Bernanke. Like Bernanke, Wheeler (a) has to herd cats and get a majority vote, not just do stuff on his own;  (b) needs to socialize what he does before he does it, lest markets flip out at what people perceive as a sudden shift (or even start panicking because they have no confidence they can predict what will happen and fear the worst as a result); while, (c) not making himself a giant target for opponents.

Bernanke’s solution was to make statements about general direction he expected the Fed to take, broad statements about the Feds priorities and goals, and rarely setting definite red lines until he actually had something lined up and could deliver. Wheeler appears to be adopting the same strategy here.

To anyone actually paying attention, Wheeler is being about as bloody transparent as his ebook. He has a list of things he thinks are important. He has set them out very clearly. He has a list of things that he thinks represent important lessons to guide the FCC’s policies. And now he is saying: “hey everybody, this ‘peering thing’ clearly falls on my list of things that are traditional problem spots in network economies. I get that. it doesn’t fit into the ‘open Internet/consumer protection bucket,’ it fits in the ‘interconnection’ bucket. So I’m on the case and we’re not ignoring this. So shift this out of ‘unthinkable the FCC will ever even consider looking at this’ to ‘we’re probably gonna check under the hood and see how this thing works.’ Not because I necessarily think anything is wrong, but because the Internet is due for its 15 year oil change and service check and even if it still drives OK, there’s this sort of buzzy sound when I drive at high speeds and the engine sometimes over-revs.”

But most people never believe this. Everyone wants ironclad definitive statements — which people don’t believe anyway so what’s the point of making them? Mind you, I understand why everyone wants definitive statements and why no one believes them. Life is uncertain and, even if Wheeler is as smart and sincere and determined as I think he is, he’s not the first guy to wade into the sausage factory that is public policy and get ground up. So all cynical and suspicious is the smart way to go. And Lord knows my facility as a professional prognosticator is about as good as any other expert going by gut rather than by data.


Let’s Say Wheeler Is Serious And Signaling Like I Say? What Then?

Happily for Chairman Wheeler, Comcast has handed him a gift by (a) buying Time Warner Cable, and (b) stirring up the biggest all-fire ruckus on this very issue before the merger review starts, rather than at the end of the merger review like last time.

Its pretty clear that Comcast has initially planned to follow its usual strategy of pretending great shock that anyone could think this merger had anything whatsoever to do with the Internet, let alone Internet peering! Shocking to suggest such a thing! Ignore the whining of our greedy competitors seeking to leverage this for unfair advantage, make it all about Netflix, isolate Netflix, transform this into an exercise about choosing sides between Comcast and Netflix, and overall work the ref so that the very act of even producing the data (which Comcast will do grudgingly, with many sighs and deletions and withholdings, forcing the FCC to decide at every stage whether to come back again for data or or just let it go) becomes regulatory overreach, making the very thought of conditions based on the data unthinkable.

I have been playing regulatory poker with Comcast for 15 years. “Trust Comcast, Comcast deals fairly, ignore America Channel/Tennis Channel/Bloomberg/BitTorrent/whatever the specific company that has actually gotten the guts to complain to the regulator despite Comcast’s well-deserved reputation for retaliation and the FCC’s well-deserved reputation for inaction. In theory there could be market efficiencies and consumer benefits and stuff and its just one company or a handful of companies and everyone else in the market is so satisfied and look how much we have improved service to customers and the customer experience” is an old, old game for Comcast. They play it exceedingly well, and it usually pays off.


But Comcast are also super smart and super nimble and know when to make a strategic concession and go all helpful. They’ve seen the blowback in the public and the press and understand that any effort to hold back the data (especially with Netflix, Cogent and L3 now out in public) will play right into the hands of merger opponents characterizing them as a well connected giant using its political leverage to prevent the DoJ (and FCC) from doing its job. Also, if Comcast keeps rolling its eyes and saying there’s no way anyone could possible imagine that this merger has anything to do with broadband, then the FCC very publicly demands the docs and gets them, it may persuade Wall St. that the deal is in trouble because the FCC and DoJ would actually be standing up to Comcast rather than rolling over, wagging their tails and asking for belly rubs.

So Comcast does its own regulatory signaling and indicates that of course they expect DoJ to do “due diligence” and look at the Comcast/Netflix deal, given the loud and totally unwarranted accusations by those parasitic ingrates at Netflix who tried to game the system by using “peering” when they aren’t peers. And since DoJ is getting the documents, the FCC will also ask for the documents. Suddenly, the “unthinkable” act of getting educated on how this works becomes thinkable. Comcast will insist on providing evidence about peering and the mechanics of the industry favorable to its position. Merger opponents will submit evidence in favor of their view of the market. Because this is an investigation, DoJ and FCC can do a lot more to compel production of relevant (and accurate) evidence.

If it looks like the DoJ and FCC don’t like what they find and would impose some serious conditions, Comcast will argue that this is an industrywide issue not fit for specific merger conditions. At which point, as with net neutrality, Comcast flips from enemy to ally in making sure the rest of the industry has to live with whatever they get stuck doing.  But even if the FCC and DoJ leave it alone for now, they will have breached the “it’s utterly unthinkable to even look at peering” barrier.

The worshippers of the gods of the marketplace will no doubt once again be horrified and bemoan the fact that agencies use merger review to “educate” themselves on the industry in this fashion at the expense of poor little companies like Comcast and TWC.  As this position is utterly inconsistent with their equally passionate argument that (a) prophylactic rules aren’t necessary because antitrust does all we could ever need, and (b) we should never make decisions without perfect data, I find their outrage as unpersuasive as it is utterly predictable.



Yes, Wheeler gets it on peering. No, he doesn’t think it is part of net neutrality. Yes, I happen to agree with this and did long before Wheeler started saying it. No, the fact that Wheeler is saying peering is interconnection, not part of net neutrality/open Internet, doesn’t mean he’s blowing it off. Yes, it will be part of the Comcast/TWC merger review. No, I have no idea whether Wheeler will decide whether the FCC will actually do anything or not. But — yes, that’s OK, because we actually should get information first before rushing to any decisions. But — no, “get information first before making decisions” does not mean “maintian willful ignorance so you can’t make any decisions.

Clear now?

Stay tuned . . .


One Comment

  1. Howard, entertaining post and pretty much spot on assessment of the forces at work. A couple of thoughts:
    1) it really IS quite simple. despite all the seeming complexity of convergence and rapid technological and demand shifts, it’s really all about the age-old tradeoff between layer 2 transport and layer 3 switching across the WAN, MAN, LAN, and even PAN. Yes Steve Jobs got interconnection in the device (PAN), which sparked the most recent digital boom. And the answer is simply this, incumbents will try to push the WAN/MAN demarc point as close to the core and away from the edge as possible. Kingsbury redux.

    But here’s the rub:
    2) the forces of 4K video content, cloud or condensed radio access (C-RAN), 2-way HD collaboration, and the internet of things (IoT) are all moving in the opposite direction. Their underlying supply and demand forces are necessitating that intelligence be distributed more towards the edge. In other words a tradeoff of layer 2 transport and capacity for layer 3 switching and storage.

    Hopefully the FCC recognizes these trends and can model them. It would be a mistake to solve yesterday’s and today’s problems when tomorrow’s are so much bigger and the payoffs enormous. It is early days, just as broadband was in 2000. But though it took 1-way short video 6 years to begin to scale in the form of YouTube (and only after we reached critical BB mass of 30% in 2005), all of these trends will come at us much more rapidly since the IP connections are so ubiquitous. But they are far, far from perfect from a price and performance perspective and we can make that happen in a market driven way. We can prove that it will be very generative for all if stakeholders in both camps can drop their false biases and preconceptions in the lower and middle layers of the stack.

Comments are closed