December 16, 2008  ·  Lessig

Fred Benenson’s got a nice piece about the WSJ piece. The most depressing part of this whole cycle was the news that the WSJ was sticking by the story.

On what basis, precisely? The charge that Obama was shifting policy was, and is, completely baseless. The charge that I had “shifted” my position was, and is, completely unsupported (and false). And the charge that Google was violating network neutrality principles has been shown (concisely by David Isenberg, one of the originals in this debate) to be just wrong — no one who understands what “network neutrality” (or what we used to call this before it was smartly marketed, “end-to-end“) is could believe that edge caching services, living in a competitive market, could raise NN concerns.

So they’re sticking by a story that’s baseless, unsupported and wrong. Sounds like we know where the Bushies have gone to work now that they’ve left the White House.

Update: So I’ve just had an email exchange with Christopher Rhodes, one of the authors of the piece. What surprised me most about the piece was that he was such a careful interviewer when we spoke, but that we didn’t really speak about the issue they charged me with — shifting — and I was surprised he didn’t ask or followup on that. Turns out he tried, writing to my assistant, but that I didn’t speak with him. My assistant didn’t know the context of our conversation, so her translation of the question didn’t flag it. My apologies to Rhodes. Had we connected, the story would have been different. The mistake in not connecting was mine, no doubt. And the mistake convinces me that at least with respect to me, the story is a misunderstanding (and not, as suggested, bad faith). Important lesson for me, no doubt. But for others: Please send emails for me to me. I read and respond to every email I get (save the spam-ish sorts). And while I can be behind, if you don’t get a response, I didn’t get it.

  • http://none.com TC

    Dr. Lessig:
    Many people see a significant difference in your position today vs 2006, as shown in your statements here.

    You say in yesterday’s post that it’s fine to charge content providers on a tiered basis, as long as “everyone has the right to the rates of the most favored nation.”

    This directly conflicts with your 2006 Congressional testimony, where you state that “the only companies that could afford to buy access… are companies like Google and Yahoo!… This restriction in competition would fundamentally weaken the growth of the Internet.”.

    Any reasonable person can clearly see that these statements suggest profoundly different directions from a policy perspective. The first, most recent, comment suggests that tiered pricing is fine as long as all players are allowed to pay identical rates per tier. The second, original, comment suggests that the very existence of tiered pricing limits competition and would “fundamentally weaken the growth of the Internet.”.

    Can you reconcile these two statements for us?

    Here’s the complete quote from your 2006 testimony:
    “And the fourth point that I don’t think anybody can really deny, the changes that are being described, not by the very reasonable people who testified in the earlier panel, but by the leaders of Verizon and the leaders of AT&T, the changes that are being described would radically reduce competition in applications and content on the Internet, radically reduce that competition because as they set up fast lanes on the Internet, the only companies that could afford to buy access to the fast lanes on the Internet are companies like Google and Yahoo! and Microsoft and the content companies that already have succeeded in the marketplace. The next generation Yahoos! and Googles cannot buy access to the fast lane, because they would face a barrier to entry that would restrict competition. This restriction in competition would fundamentally weaken the growth of the Internet.

  • http://blogan.net Brent Logan

    Why the snark at the end? Is everything partisan?

  • http://www.greenscreencinema.com Green Screen Cinema

    Professor Lessig, I again ask you to clarify your position on one network provider charging both parties in an internet transmission (when two network segments exist). This is the “plot” that the WSJ clearly made up, but your refusal to address this has made me wonder if your MFN requirement actually allows this.

    Your statement “The regulation I call for is a “MFN” requirement — that everyone has the right to the rates of the most favored nation.” suggest that network providers could charge both parties in a transmission as long as those charges were equitably distributed. Since Google pays for its own bandwidth (up to their provider’s backbone), why would AT&T ever be able to charge them anything without a violation of network neutrality principals? If a content provider were at any time required to pay a fee to the consumer’s network provider it would give them market advantage over content providers that could afford no more than their own bandwidth fees.

    The Google edge caching proposal clearly does not interfere with network neutrality, I think we can all agree on that. The two concepts of edge caching and MFN are orthogonal: only technically competent/complimentary companies with sufficient bandwidth needs should be allowed into the AT&T datacenter to edge cache, so most “nations” would be excluded from this kind of arrangement. But if consumer charges are the same, it would have no impact on the neutral routing of data.

    Your application of MFN to the Google proposal raises the issue, however. Yes, the WSJ is nefarious in it’s depiction of Google, but I would like to hear how you look upon the “both sides pay” scenario that WSJ envisioned. Is there any valid reason for a company that holds the consumer captive (AT&T) to be able to charge content providers, for anything? I would argue that there is no time at which traffic should be double charged; I would argue that what we really need is “routing neutrality”. This would prevent content companies and network providers from conspiring together, a nerve which the WSJ has hit upon.

  • szlevi

    “Brent Logan:
    Why the snark at the end? Is everything partisan?”

    Oh, pleahhhse.
    We’ve had more than enough of this ind of “innocent” (fake) ‘non-partisan’ question-line BS in the press for a decade – time to call a spade a spade, period.

  • lessig

    @TC: Let me limit the response today to whether what I said in April is different from in 2006. Thursday I hope to have a second to argue why I think the rule I promoted in April is better than the rule I criticized in April.

    The concern that has been at the center here is a concern about leveraging. My objective has been to find the least intrusive regulation necessary to avoid the incentive to leverage (using power in one market to extract rents in another). Chicago school types say leveraging isn’t possible. I don’t buy that (for the reasons van Schewick outlined). But I do think skepticism about when it is possible is appropriate, and I have had that skepticism for ever.

    So from the perspective of access-tiering — my question has always been, when will leveraging make sense. And the concern I had in 2006 (and today) is when giving network providers the right to condition access might create the incentive to leverage. One case was when the provider could exclusively condition access to a service (e.g., a fast lane). In that case, the incentives to strategic leveraging are greater than when exclusivity isn’t possible. So, e.g., when I said “Those competitors will never have the opportunity to
    compete effectively against Google Video if Google Video can buy the fast lane.” I really meant “buy the fast lane.” The scenario I was worried about was an exclusive deal that blocked others from an essential service.

    Now I’m the first to admit that market power may create enough of a difference such that even with just a MFN regulation, that may not be enough to create the incentive for broadband providers to select a business model of abundance rather than scarcity. That’s the point of my FCC testimony — “shock and awe” isn’t necessary; a step by step model of regulation within a clear regulatory framework is sufficient. So the FCC should still (even under the relatively minimal rule I advance) watch to make sure the right kind of competition has developed, and that pricing isn’t being used to indirectly effect leveraging. But you’ve got to put regulation in the context of assumptions about how competition will work. And a MFN, in my view, still feeds competition.

    @Green Screen: Again, there are relatively few contexts in which competitors can “double charge.” Or put differently, whatever is being charged is whatever could be charged — lots of times. Sometimes that’s not true. That’s my concern about leveraging. But when leveraging isn’t rational, I’m not worried about double charging.

  • Adrian Lopez

    “Again, there are relatively few contexts in which competitors can ‘double charge.’”

    Content Producer X wants a high-speed path to its customers. ISP A has both the infrastructure and capacity in place to deliver content at those speeds but decides it wants more money. If Content Producer X refuses to pay money to ISP A, ISP A will throttle Content Producer X’s packets in such a way that Content Producer X’s customers will suffer degraded performance relative to the content’s requirements. Worse yet, ISP’s B, C, D, E and F also want their cut. ISPs A, B, C, D and E all want money from Content Producer X in addition to the money they make from home and business users, double-billing in exchange for an artificial benefit (non-throttled bandwidth).

    It seems to me that, perhaps in an attempt to make your proposal seem more fair to free-market interests, you are in fact promoting a kind of token regulation that addresses a technical legal issue without actually fixing any of the problems the law is meant to fix.

  • lessig

    @Adrian: “free-market interests”? The market creates a set of constraints. The law is crafted in light of those constraints. We all have to draw upon our understanding of the market constraints, based on an understanding of economics. Economic theory teaches that sometimes leveraging is profitable, sometimes not. The Chicago School is famous for extreme skepticism. The extremism is not merited, but some skepticism is totally appropriate. If government regulated assumed leveraging was always possible (as your post suggests), that would be a terrible mistake.

  • Adrian Lopez

    I’m neither a lawyer nor an economist, so my stance may be more practical than technical.

    Anyway, I think the bigger mistake is to think that allowing content providers to strike deals with remote ISPs won’t lead to a situation where those who can afford to strike deals get all the benefits of an exclusive deal without the deal actually being exclusive. For instance, while Google could easily afford to strike deals with the myriad ISPs that would otherwise throttle Google’s packets, iFilm perhaps could not. I’m certain that ISPs would want to charge as much as possible for non-throttled routes, such that content providers that can afford to pay the most, like Google, would effectively set the price below which everybody else is excluded from non-crippled routes.

    Net Neutrality legislation should be crafted in a manner that protects the traditional relationships between ISPs and their clients: the one that provides the physical connection is the one that gets to price your bandwidth. ISP X can charge its customer, Content Producer X, whatever it wants, but ISPs A, B, C, D and E — who have no existing business relationship with Content Producer X — should not be allowed to cripple Content Producer X’s packets.

    Any law that doesn’t address that is, in my view, seriously flawed.

  • Paulo

    No wonder the Wall Street Journal thought your position had shifted Professor, it shifts by the day.

    Today:

    “no one who understands what “network neutrality” is could believe that edge caching services, living in a competitive market, could raise NN concerns.”

    You should tell FreePress. The argument for Neutrality regulation is that the US does not have a competive market:

    http://www.savetheinternet.com/=lie4

    Big Lie of the Week: No. 4

    “There simply is not enough competition between different technologies to produce any kind of deterrent to discrimination.”

    So Neutrality regulation is needed because the market isn’t competive, but the market is competitive, so Google’s fast lane doesn’t violate Net Neutrality.

    Kinda selective, aren’t you?

  • Paulo

    No wonder the Wall Street Journal thought your position had shifted Professor, it shifts by the day.

    Today:

    “no one who understands what “network neutrality” is could believe that edge caching services, living in a competitive market, could raise NN concerns.”

    You should tell FreePress. The argument for Neutrality regulation is that the US does not have a competive market:

    http://www.savetheinternet.com/=lie4

    Big Lie of the Week: No. 4

    “There simply is not enough competition between different technologies to produce any kind of deterrent to discrimination.”

    So Neutrality regulation is needed because the market isn’t competive, but the market is competitive, so Google’s fast lane doesn’t violate Net Neutrality.

    Kinda selective, aren’t you?

  • http://managingmiracles.blogspot.com/ Steve

    @Paulo

    Enough sniping (and double-posting). Last-mile vs. backbone/core market.

  • Paulo

    (Apologies for the double post)

    @ Steve, you need to bone up on competition law.

    Google is a vertically integrated content provider. Google is offering a Fast Lane for a fee to ISPs to carry its content. How long do you think your ISP would survive in the market if it offered a) no YouTube at all or b) YouTube slower than rival ISPs who pay the Google Fast Lane access charge?

    How do you feel about “We deliver slower YouTube than anyone else” as a marketing slogan? And how about “Our Google searches are 50 per cent slower!”

    Let me know when the penny drops.

  • lessig

    @Paulo: My surprise was not that the WSJ misunderstood me. I’m used to what I said not being understood by even brilliant sorts (I read my own exams…), and I can see how in this context the 2006 testimony suggests something different. My surprise was that the Journal didn’t speak to me about the “shift” — and that surprise has been eliminated by the account above of not connecting. So no surprise on my part here, now, at all.

    But what’s not true is that my position has changed. Here’s the best evidence I’ve got. In this paper by Scott Hemphill (not a supporter of NN regulation), Scott writes as follows:

    Not all network neutrality proponents insist upon a zero-price rule. Others, including Lawrence Lessig[31] and a senior Google executive (dissenting from the firm’s official position),[32] appear to accept access fees imposed upon content providers, provided that an offer made to one content provider is also extended to its rivals.[33]

    The note (31) is as follows:

    31 See, e.g., Lessig AEI Presentation, supra note 15, at 1:04:20 (answering, in response to question, that charging Google and rival the same amount would not raise application tiering issues); id. at 58:30 (stating that only concern is viewpoint discrimination); Lawrence Lessig, Lecture at Center for American Progress, The Withering of the Net: How D.C. Pathologies Are Undermining the Growth and Wealth of the Net 14 (June 16, 2006) (transcript available at link) (accepting practice where access provider charges rival content providers same premium price for video transport).

    I’ve not been able to listen to the AEI presentation. But the text quoted from the CAP presentation is as follows:

    The kind of neutrality regulations I’m
    talking about would be restrictions on the kind of business proposals that network owners are allowed to offer people in the world. So they’re not allowed to offer Google a special price over Yahoo! for the same kind of service. Well it’s pretty easy to police that. Of course, they might try to do it secretly. They might try to have these secret deals where
    they strike and there’s a problem with internal versus external content, no doubt, but the point is what regulation should be trying to do here is to create enough incentives for network owners to build a business model that doesn’t depend upon discrimination.

    This event was June, 2006. The AEI event was before it. That was my view then. It remains my view now.

  • http://www.greenscreencinema.com GreenScreenCinema

    Thanks for the response Professor Lessig, I truly appreciate it.

    What I think you are saying is that other than when a network provider forces additional charges on a content provider (and has the leverage to do this) you are not concerned about double charging because a free market would simply not tolerate it. I think this kind of free market thinking, however, turns a blind eye to the possibility that a wealthy content provider could freely offer to pay a surcharge, as it seems Google has done, in order to create a barrier to entry for others.

    The fees Google is offering to pay for edge services may simply be the Trojan Horse that Google is using to limit competition in the online video space. If Google can make this surcharge the standard cost of business, it will keep out those that cannot pay it. In a truly neutral environment a content provider would only need to pay for their connection to the backbone, and could freely choose any ISP to perform this service. In an edge caching world the content provider would have to pay whatever fee the consumers’ network provider decided to charge for the edge services. And furthermore, by setting the standard for edge services Google will keep its own integration costs down while causing even more disruption for competitors.

    The position I would like you to consider is that by forcing network providers to maintain the same latency for all traffic (regardless of type) and disallowing special treatment such as local edge servers the “edge services barrier to entry” could not be constructed. Why should placing content one hop closer to the consumer really make any difference anyway (in a neutral world)? The same content could live just outside the network provider’s network, and there it would be accessed through standard routing protocols. As soon as it jumps that barrier it is no longer “neutral” in any sense. The network provider can exercise load balancing protocols, availability tests, etc. to determine how to route traffic to the edge servers. This is inherently discriminatory.

    I hope you will consider not only the importance of standard pricing, but of standard routing protocols. This would help to prevent any future barrier to entry the online video market.

  • http://www.socialsecuritybullshit.com Steve Baba

    Paulo: The double post is Lessig’s fault for sticking with a web host so bad that it takes minutes to post, making many people double post by mistake. Ironically, this example provides a counterexample to your how long would a company last if it provided bad service question. And there are thousands of web hosts while only 2 to 5 broadband providers.

  • http://bennett.com/blog Richard Bennett

    According to Lessig, he’s been arguing for net neutrality (the belief that specific regulations flow from the structure of the Internet’s unremarkable end-to-end layer) since 1999. What’s more disturbing, a guy who has no particular knowledge of networks taking a position on how to regulate them and then refusing to change it for nine years, or a guy whose views evolve as he learns the subject matter?

    If Lessig is really this intransigent, he’s more like W. than all but a very, very small number of people in this world. All the other people who’ve been debating this subject freely admit to changing their views on particular details, including David Isenberg.

    Lessig is the king of outliers.

  • Eddie

    Are we beginning to see the Murdoch effect gradually taking place in Editorial (top-down) into the WSJ?

  • http://sethf.com/ Seth Finkelstein

    Oh, c’mon Richard. You’re taking a cheap shot. The WSJ article was wrong about his supposed change. Why do you want to turn that into some sort of can’t-win-for-losing slam?

  • Jim Feeley

    So now that Christopher Rhodes appears to agree that his article contains a mistake (right?), will the WSJ issue a correction or clarification?

  • http://www.socialsecuritybullshit.com Steve Baba

    Don’t all politicians change their views as they progress from long-shot candidate appealing to extremists with, “heated campaign rhetoric,” to battling for the middle ground in the general election, to actually having to do something practical about the problem and pay for the solution when elected?

  • http://stryder.com Ross

    Why did the WSJ journalist accept a followup answer from your staff as canonical, and further expect your staff to understand the complete context of your one-on-one interview. Sounds like bad faith. I don’t accept your apology. WSJ staff should know how to use email corectly without you having to explain it. Not your fault here.

  • http://managingmiracles.blogspot.com/ Steve

    @Paulo

    Your initial characterization of the two quotes was disingenuous. The argument was clear.

    Now, if you disagree with that argument and wish to make a speculative and novel anticompetitive leveraging argument, go ahead. Just be genuine about the position of the other side.

    FWIW, I have no problem with the “ESPN 360” approach. I just think it’s an idiotic business model that Google seems unlikely to attempt. Given the current market structure, it certainly isn’t a Sherman Act violation.

  • eksabajt

    Prof. Lessig:

    I’m keen to hear your response to GreenScreen’s second post (and Adrian’s as well). It seems to be a valid concern that large content providers would be able to price out the competition despite the MFN requirement. What do you think of GreenScreen’s suggested requirement of standard routing protocols?

  • Serge

    The trouble with the reaction to the WSJ article is all the vitriol. It may, indeed, be that Larry Lessig’s version of net neutrality always allowed the rich to pay for better access. But the we’re-all-equal rhetoric surrounding it has long obscured that feature, to the extent it was always present — and many of the statements of a great many net neutrality advocates go in the opposite direction, claiming that the whole point of net neutrality is that the rich and poor are on an equal footing.

    The vitriol against the WSJ is improperly directed. Rather, it’s part of a vigorous debate that needs to happen between net neutrality advocates who, in fact, are advocating different things.

  • Griffin Drutchas

    Serge has summarized well my recent realization in learning about network neutrality — neither the idea nor the popular support for it is monolithic. Serge’s post, along with the evidence presented in Lessig’s response to Paulo on 17 Dec, have (for the time being) convinced me that those vitriolically accusing Lessig of changing position are gripped by the angst of confusion in finding Lessig to be more of an economic Liberal than a Comrade on NN.

    Fortunately, Christopher Rhodes’s carelessness is our gain. At least for me, the debate surrounding Lessig’s consistency has motivated deeper thinking on NN. Whether or not one accepts Lessig’s self-defense (And one should, considering it rests on documented evidence absent from the claims of his critic), one need look no further than C. Scott Hemphill’s NN paper to understand that NN cannot be equated with a uniform price doctrine.

    In the struggle for the soul of the internet, I agree with Serge that NN advocates must engage in a vigorous debate to discover any internal schisms before they undermine the opportunity to preserve and promote the Internet as a means of social production and cultural evolution.

  • Adrian Lopez

    “Whether or not one accepts Lessig’s self-defense (And one should, considering it rests on documented evidence absent from the claims of his critic), one need look no further than C. Scott Hemphill’s NN paper to understand that NN cannot be equated with a uniform price doctrine.”

    Whether network neutrality can be equated with a uniform price doctrine depends on how you define network neutrality in the first place. We’re talking about policy here, not natural law. The policy is what we make it.

    I consider a form of extorsion to cripple a content provider’s packets for failure to pay an ISP with which the content provider has no existing business relationship. Prohibiting extorsion is good policy.