July 2, 2015

avatar

Christopher Yoo on Comcast and Competition: When Antitrust Lawyers Do the Math

Update:

Christopher Yoo gave his talk, and I encourage you to watch it. As you can see from the rather extensive comment thread below, Yoo does not think that my critiques are fair, and he is more than a little bit upset that I trolled him. Nevertheless, upon review of the debate, I believe that you will find that the TL;DR is:

  • He admits that his “well established” means of quantifying broadband competition are anything but.
  • When I ask him to verify basic assertions that he made in answering my critique (eg: Netflix has historically paid ISPs for “carriage”), he dodges and claims that I don’t understand the industry.
  • He thinks that ISPs are incapable of traffic shaping that they were already doing six years ago.
  • He admits that ISP discrimination, which has recently helped ISPs to negotiate “paid peering” (a.k.a. reversing the transit relationship), does in fact destroy the “bill-and-keep” model that has historically been de facto for broadband service, and that this discrimination leads to a terminating access problem.
  • He claims that, nevertheless, last-mile market power does not exist when “networks at the core of the network engage in settlement-free peering.” I’m not sure why.
  • He avoids answering my basic critique of his legal interpretation of Time Warner v. FCC, 240 F.3d 1126, which is key to his “viability” standard.
  • He appears to feel that the government should make regulatory decisions based on what feels equitable or at least some economist’s definition of welfare-maximization (presumably a leading-edge neo-schumpeterian n-sided-market economist).
  • He accuses me of not reading his scholarship, even after I quote from it liberally, and equates ex-post enforcement of antitrust principles to “regulatory intervention” (I suppose we could have a semantic debate about this one, but the chasm between concrete rules and his notion of antitrust is great.)
  • He’s upset that I keep mentioning his own public disclosures of corporate funding. My view is that if you are a decent academic, the degree to which corporate support is relevant is indirectly proportional to the merit of your scholarship. This calculation is an exercise left to the reader.

Today at 12:30, Christopher Yoo will give a live-streamed talk at CITP entitled “The Open Internet in the Aftermath of Verizon v. FCC: What Comes Next?” Yoo will talk about the Verizon v. FCC ruling that overturned the FCC’s network neutrality rules, and place them in the context of the proposed merger between Comcast and Time Warner. Yesterday, unnamed sources within the FCC gave a possible answer to Yoo’s question: a “fast lane” for sites and services that pay for preferential access to Comcast’s customers. The FCC is reportedly considering new rules that would permit broadband Internet providers to discriminate against specific content as long as they don’t do so in an “anticompetitive manner.” The FCC will then be left to decide what makes for anticompetitive behavior — a domain typically left to antitrust law.

This FCC decision about network neutrality is taking place in the context of the controversial proposed merger between Comcast and Time Warner — which itself is undergoing antitrust review. The core question is whether or not competition will be harmed in a way that is bad for consumers, or for business on the Internet in general. One of the most-touted examples of alleged network neutrality violations (or of interconnection discrimination, depending on how you look at it) is Netflix. Netflix recently announced plans to raise prices, noting Comcast’s success in demanding payment from the backbone network providers that deliver Netflix content to paying Comcast customers when they request it. Traditionally, “last-mile” providers like Comcast would pay backbone providers to connect their customers to the rest of the Internet, but in recent years Comcast has become so large that it has managed to reverse this relationship. They have done so by threatening to cut off or degrade connections when particular backbone providers begin to deliver higher volumes of Comcast customers’ bandwidth requests. Companies like Netflix argue that they are left with nowhere to go, because Comcast will take a similar approach with any backbone provider they use.

David Cohen, Comcast’s Executive Vice President, is a University of Pennsylvania law graduate and has been arguing his case in the language of antitrust law. Serving as “Comcast’s Real Repairman”, he takes the following approach when lobbying:

Mr. Cohen approaches his advocacy for Comcast’s transactions like the litigator he once was. “It’s the same way you build a case. You understand all the facts down to a minute level,” he says. “And then you pull the lens back so everyone else can understand it, and organize the advocacy in a way that is compelling and persuasive.”

He makes several points about the merger with Time Warner Cable: The two companies do not compete in the same cable markets, so the deal in his view will have no impact on consumers.

I can’t speak to the veracity of the claims reported by the Wall Street Journal, the analysis of the New York Times, or the explanation by the Washington Post, but I do think that it will be interesting to see what Yoo will say. I have a pretty good idea. He recently testified before Congress, alongside cable lobbyists, stating that:

In short, established principles of antitrust and communications law dictate that the merger is unlikely to harm consumers in either market. In fact, technological and economic changes are transforming the markets in ways that should make the prospect of anticompetitive harms even more remote.

Yoo has been giving his opinion on the state of broadband competition for many years. In his book, published by free-market think-tank AEI, he wrote that:

Any change necessarily creates new winners and losers, which will inevitably prompt those disadvantaged by a change to argue that it is anticompetitive. […] Far from being regarded as a problem in need of remediation, change should be embraced as a natural and indispensable part of the process of the Internet’s evolution.

Yoo is an antitrust lawyer, so for him everything has to do with whether a given market is competitive. Back in 2008, I wrote the following about his analysis of the home broadband market:

Yoo admits that the last-mile market is concentrated. “Were network neutrality designed to promote competition on the side of the market in which last-mile providers meet end users, the market would be local in scope and sufficiently concentrated to provide an arguable basis for regulatory intervention” (Yoo, “What Can Antitrust Contribute to the Network Neutrality Debate?“, 2007). However, he attempts to define the market nationally, claiming that the relevant scope is the field of all national broadband providers.

I went on to describe why the local market seemed to be the most critical market in competition analysis. I argued that what matters is how many options a given consumer has, not how many options the superset of all consumers have collectively. Providers that do not serve a particular home, I said, are not competitors for that consumer despite Yoo’s reasoning. “If his reasoning held, one might just as well include all international broadband providers as well, making for a very rosy picture indeed.”

On Yoo’s telling, antitrust law would have us adjust our count of the two home broadband providers that most consumers face. On his accounting, this might be closer to two hundred or two thousand. It’s hard to do the math.

Comments

  1. avatar Christopher Yoo says:

    As I have emphasized in my published work, broadband Internet access providers operate in two distinct markets, each with a different geographic scope. Most importantly here, the merger would neither increase nor decrease the level of concentration in this market at all, as Comcast’s service area does not overlap with Time Warner Cable’s. Consumers would have the same number of options in broadband Internet access providers the day after the merger that they did the day before.

    Broadband Internet access providers also serve the market in which they sell the connections that content, application, and service providers (sometimes called edge providers) use to obtain access to end users. The geographic scope of this market is national. Consider the following hypothetical example: Smartphone manufacturers care about the total number of phones that they sell and not about whether they are sold in any particular market. For them, it is national reach, not local reach, that matters. In this market, a mobile network operator that controlled thirty percent of the national market would be unlikely to be able to harm the phone manufacturer, as they would still have an open field of 70% of the market. The same is true for edge providers, whose viability is similarly the result of their overall scale on a national or international basis and the national scale of the people with whom they bargain (although the interchangeability in each market must be analyzed carefully). Because the national market share of the merged company would be roughly 30%, the same logic applies.

    All of this represents well established antitrust principles regarding the importance of carefully defining geographic markets as well as product markets. It is also laid out painstakingly in my work.

    Lastly, I must register some disappointment in the tone of your post. Chiding antitrust lawyers for their inability to do math is basically an ad hominem attack that does not engage the merits of my underlying argument (and misunderstands my argument at that). It also does not promote a better understanding of the issues either by the people engaged in the debate or the audience reading the debate. For example, if I were to fire back by chiding engineers for their inability to read, it would obscure the debate rather than provide enlightenment. I’m engaging you in this round in the hope that Freedom to Tinker is intended to be a place for serious discourse about issues. If it is a venue for mischaracterizing other people’s viewpoints and engaging in rhetorical ploys that are designed to appeal to those who already are convinced by one side of the argument, please do not take my future silence as any acquiescence in any aspersions cast in my direction.

    • avatar Steve Schultze says:

      Chris,

      There is no ad hominem here. I understand your work, and your version of the antitrust principles you espouse. I’m sorry if my joking about doing math fell flat — I am sure that your arithmetic is impeccable.

      I simply think that, taken on their face, your arguments don’t add up (oops, I did it again… not a dig at your accounting skills!).

      Why not define the market globally?

      Why does increased concentration in last-mile providers not further increase their leverage in the interconnection and global content market?

      Why should we think that the market power of of edge providers that are locally duopolistic balances out with negotiating power of the competitive backbone or content providers in such a way that does not lead to market failure? The questioners at the end of your talk did a good job of probing you about how your home-buying analogy is inapt, and I remain unpersuaded by your response.

      It’s not inaccurate to refer to two different markets, but they are by no means distinct. Was it a Freudian slip that you initially commingled them before changing your description below? Market power in the last mile translates into market power in the interconnection and content markets, despite your attempt to distinguish the dynamics at hand from the terminating access monopoly.

      If your argument is that the local broadband market is indeed failing, and that this is a problem, but that any regulation is far worse, then just be forthright.

      If your argument is that antitrust perpetually defers to any market, no matter how dysfunctional, then the answer to the question of “What Can Antitrust Contribute to the Network Neutrality Debate?” is quite simply, “nothing.”

      • avatar Christopher Yoo says:

        Thank you for engaging me on the merits. Here are my quick responses:

        It is possible that the relevant market might be global instead of national. Everything depends on substitution patterns for the complementary service. For example, with respect to equipment, I suspect the market is global. For content and services, language differences and the desires of advertisers make it likely that the relevant market is national, although I could see the demand within a country for foreign language materials and scale economies may provide some spillovers that could push towards a supranational market definition.

        Regarding the bargaining power of terminating providers, as I stated during my CITP presentation, terminating providers only have bargaining power under pricing regimes built around calling-party-pays or sending-party-pays. It does no arise under peering or bill and keep. (Note also that when it does arise, it does not matter whether or no the terminating provider has market power.) Whether concentration in the last-mile market increases market power depends on market structure. If a merger increased market share for 10% to 15%, no one would regard it as problematic. If a merger increased market share for 50% to 100%, everyone would regard it as problematic.

        The bargaining power of last-mile providers is limited by several complex factors. First, the bargains are not all or nothing. Multihoming and the existence of multiple paths limits the bargaining power of the last-mile providers in any particular negotiation. In addition, the relevant question is relative bargaining power. One of the most interesting recent developments is the way that Neflix was able to make pretty much every ISP around the world join Open Connect and terminate Netflix traffic for free. Many of these last-mile providers have much stronger positions in their local markets than any U.S. provider. Their market position in the last mile didn’t save them. Don’t get me wrong: Netflix obtained its bargaining power by investing billions in access to content, and it deserves to enjoy whatever leverage it derives from its willingness to take risk. It just shows how you have to actually do the analysis and not just demonize last mile providers. Put in terms of the other market at issue in the Comcast-Time Warner Cable merger (multichannel video), the moves in prices during recent years makes it clear that the content providers have the upper hand, not the last-mile providers.

        No Freudian slip, by the way. When I cut and pasted text from Word, it was truncated, and I did some quick surgery before realizing the edits were too quick. Simply put, market power in the last mile may or may not translate into market power in the interconnection markets. You have to look at substitution patterns, define the relevant markets, and then analyze the industry structure. If you don’t believe me, ask the last mile providers who agreed to terminate Netflix traffic via Open Connect for free when they used to be paid transit for doing so.

        No antitrust scholar worth her or his salt believes that antitrust perpetually defers to markets. Modern antitrust (that is antitrust of the last 50 years or so) understands that antitrust is a fact-intensive inquiry that looks for empirically based reasons and evaluates price effects and the relevance of scale economies and other efficiencies in the context of specific models without indulging in generalized concerns that big is bad. Changes in bargaining power are evaluated in a specific factual context and not simply asserted. Those who forego this analysis risk falling into the symmetrical pitfall of reflexively thinking that all mergers are bad without doing the analysis.

        • avatar Steve Schultze says:

          All right, let’s dive into your latest barrage of assertions.

          First: you claim that locally concentrated ISPs do not cause harm because they operate under “bill-and-keep.” As you no doubt know, bill-and-keep is a regulatory regime in which the government mandates that each network foot the bill for the costs of its own customers. Further down this thread, Mitch Golden points to the excellent Vox.com article, “Comcast is destroying the principle that makes a competitive internet possible.” That principle? Bill-and-keep. Last-mile discrimination (in the form of source discrimination or interconnection arbitrage) violates bill-and-keep, so you are actually describing not the preservation but rather the evisceration of bill-and-keep. This definition is of course consistent with how Phil Weiser explains bill-and-keep in his touchstone co-authored book on the telecom law. You thank Weiser in your opening footnotes to “Beyond Network Neutrality,” right before you caveat that, “after this article was substantially complete, I was retained by the National Cable and Telecommunications Association (“NCTA”) to consult on matters related to issues discussed in this Article.”

          Bill-and-keep, because it does not exist in this case, will not save us.

          You also mention that terminating access leverage “does [no] arise under peering.” I presume that you are not trying to argue that the last-mile market traditionally involves peering relationships, but if so I am happy to digress into the difference between peering and transit–and the relevant undisputed market definitions. In any case, I will refer you again to that excellent Vox explainer. The audacity is that the non-locally-competitive last-mile providers have begun to demand that the money flow in the opposite direction from normal (contrary even to a no-fee bill-and-keep structure). Only the most concentrated last-mile providers have succeeded, which is evidence of precisely how last-mile consolidation allows non-optimal leveraging of terminating access market power once consolidation reaches a certain point (not evidence of something else that would give pause to Netflix bargaining power… which you gestured toward in your talk but did not actually explain).

          We now turn to the “complex factors” that you claim will obfuscate our ability to understand last-mile bargaining power. Multihoming and multiple paths are irrelevant in a market in which the last-mile provider can discriminate regardless of source IP or BGP route. This was a market reality years ago, and even telecoms in developing nations have obtained the ability to do this. Or perhaps you are trying to regurgitate your telecom lobbyist peer’s bald assertions to this effect. The lobbyists seem to say, with a straight face, that because of the diversity of links into their last-mile networks, they can’t possibly discriminate based on source.

          One of the most interesting recent developments is the way that Neflix was able to make pretty much every ISP around the world join Open Connect and terminate Netflix traffic for free. Many of these last-mile providers have much stronger positions in their local markets than any U.S. provider. Their market position in the last mile didn’t save them.

          What price were those ISPs obtaining in exchange for “terminating” requests made by their customers before OpenConnect? It’s not nothing… it’s negative. U.S. last-mile providers have reversed the payment structure as they have consolidated, whereas competition and governmental checks on market concentration have counterbalanced these dynamics in much of the rest of the world. What’s your calculation of the HHI of Comcast in the local market? Of course, I’m sure that you will reply with a different market definition that includes enough entities that the HHI is nothing to worry about. I don’t think that this will be much comfort to Dr. Camp, who described in her comment below how she has given up her initial optimism in light of her personal experience and engineering expertise.

          You say: “ask the last mile providers who agreed to terminate Netflix traffic via Open Connect for free when they used to be paid transit for doing so.” I have no idea what you are talking about. Who? Used? To? Get? Paid? To? “Terminate”? Netflix? Traffic?

          Allow me to digress a moment to your mention of the edge provider’s ability to remain viable as evidence that the market is working just fine. In your original “What does Antitrust Have to Offer” paper, you claim that so long as edge (content) providers “remain viable,” antitrust law tells us that concerns are beside the point. You cite Time Warner v. FCC, 240 F.3d 1126, (D.C. Cir. 2001), for this “not dead yet” test, but in that case the court considers viability “independent of concerns over anticompetitive conduct” (p. 1133). It wasn’t a factor in the antitrust analysis! Again, this is straightforward misdirection and obfuscation.

          More to the point, we have learned now (from you) that even in the “up is down” potempkin village that you have constructed out of bits of antitrust detritus, the “national scope” that I criticized in my initial post is indeed quite pliable. This is because despite your subsequent claim that “well established antitrust principles regarding the importance of carefully defining geographic markets as well as product markets”, the best you can muster is that ISPs also compete in a non-local market (the death of bill-and-keep notwithstanding) and that the best approximation of this market is national because of language barriers. This language barrier argument is a new one on me (I don’t see it laid out painstakingly in your work from which I am quoting), and such a naively isolationist view of the English-speaking world is unpersuasive. But perhaps I am just having illusions of a borderless world.

          No antitrust scholar worth her or his salt believes that antitrust perpetually defers to markets. Modern antitrust (that is antitrust of the last 50 years or so) understands that antitrust is a fact-intensive inquiry that looks for empirically based reasons and evaluates price effects and the relevance of scale economies and other efficiencies in the context of specific models without indulging in generalized concerns that big is bad. Changes in bargaining power are evaluated in a specific factual context and not simply asserted. Those who forego this analysis risk falling into the symmetrical pitfall of reflexively thinking that all mergers are bad without doing the analysis.

          Please provide a cite from your work on telecom that demonstrates something other than market deference.

          I do agree with something that you said in your article, “The Changing Patterns of Internet Usage:”

          Instead of creating regulations that lock in any particular vision of the network’s architecture, policymakers should create regulatory structures that give industry actors the latitude they need to experiment with different solutions.

          Whereas that work was “made possible by a research stipend from the Time Warner Cable Research Program on Digital Communications,” my critique is made possible by some free time I had over the weekend.

          The corollary to your point is that when industry actors experiment with clearly anticompetitive behavior that stifles innovation in the most innovative parts of the network, policymakers must reconsider the regulatory structures that led to such a state of affairs.

          • avatar Christopher Yoo says:

            My comments about peering/bill and keep were in response to the oft asserted belief that the terminating access monopoly always gives companies like Comcast an ability to harm competition. In fact, that was what you suggested when you said, “Market power in the last mile translates into market power in the interconnection and content markets, despite your attempt to distinguish the dynamics at hand from the terminating access monopoly.” In fact, market power in the last mile does not always translate into market power in the interconnection and content markets. First, control over termination only translates into market power over interconnection and content markets only under certain pricing regimes, particularly sending/calling party pays. Most importantly for the Internet, it does not exist under a world in which networks at the core of the network engage in settlement-free peering. Second, this problem does not even require a monopoly. It arises even when last-mile providers face competition. It is the calling/sending party pays pricing regime, not the market structure, that is the source of the problem.

            My comments were an attempt to add nuance to the type of blanket statements like the one you offered, not to go through the all ins and outs of paid peering. The current literature indicates that paid interconnection agreements like the one between Comcast and Netflix have both upsides and downsides for consumers. On the downside, as the Vox explainer correctly points out, a general shift to paid peering would cause the terminating access problem to arise in the Internet. There are several other considerations on the upside. First, paid interconnection arrangements like the one between Comcast and Netflix are considered more equitable in that they place the burden of the increased traffic caused by the growth of Netflix on Netflix users without making non-Netflix users bear the burden of increasing costs that they did not create. Second, the economics of two-sided markets, which is one of the hot topics these days, indicate that as a theoretical matter, the welfare implications of paid peering are ambiguous. When that theory is parameterized with some basic data (such as the fact that advertising revenue flows into the network on the content side and the costs are higher on the end-user distribution side of the interconnection points), it becomes likely that paid peering increases welfare.

            Contrary to your suggestions, multihoming and more complex architectures do have a fairly significant impact on the economics. When there is only one way into a network, failure to reach an agreement disconnects end users, and the last-mile provider enjoys the leverage of the full value of the connection. When there are multiple ways into a network, the last-mile provider’s leverage is limited to the difference between what it wants to charge and the cost of the next best path into the network. Comcast, for example, maintains 40 peering relationships and more than 8000 transit relationships. A network that is unable to reach an agreement simply has to incur the cost of routing through another connection, which given the dropping prices in core transit these days is likely to be slight. A company that wanted to penalize this traffic would have to sniff all 8000+ connections and separate traffic originating from the party against which it was trying to discriminate from all of the other traffic. Even if this remains a concern, the fix is more straightforward than having the FCC regulate interconnection more generally.

            Regarding the viability of edge providers, if one examines the value chain, they are capturing the lion’s share of the surplus being generated by the Internet. Google, Apple, and Microsoft rank among the four largest companies in the world. Indeed, depending on the day, the market capitalization of a single edge provider often exceeds the market capitalization of the entire cable industry combined. What’s more, they have a geographic footprint that is national (if not international, as we discussed earlier), while last-mile providers typically only reach 30% of the nation.

            Also, if you think that Netflix isn’t using its market position to force ISPs to terminate traffic for free for which they used to have to pay (or at least have included in ratio balance of a peering contract, which is the barter way of making payment), you simply don’t understand the industry.

            Regarding your questions about financial support for my research, I would suggest that it is no different from other academic institutions (let alone advocacy organizations). For example, CITP has a Microsoft fellow and discloses that it receives funding from unlisted private sources that probably directly or indirectly paid your salary at one point. As long as CITP insists on academic independence (as every serious academic program does, including mine), I have no problem with that.

            Rather than going back and forth about these and the other points you made, I will stop here and say that this will be my last post. First, you responded to my one-page comment with a two-page post that raised a wide range of other issues. Instead of narrowing the issues, you are making them wider.

            Second, despite your attempt to pass off your crack about lawyers not being able to do math as a joke, the tone of posts continue to presume my bad faith, as evidenced by your use of the phrases “barrage of assertions,” “regurgitate,” and the accusatory tone in almost every sentence of your post. You are not interested in understanding what I am trying to say or engaging in a dialogue. You are trying to score debating points in a snarky way that is convincing only to people who are already convinced rather than enlightening people. I expect such tactics in political circles, but I thought that as a university-based venue, Freedom to Tinker was about exchanging ideas, not ridicule.

            Lastly, you ask that I point to where I have advocated something other than market deference. Simply put, your request shows that you still haven’t bothered to read my work fairly. Every article of mine that you cite advocates regulatory intervention. In fact, I have been criticized in some quarters for doing so, but it seems to me the more analytically sound position to take. The difference is that I advocate an ex post case-by-case approach instead of an ex ante prescriptive approach. Those are two different approaches to regulation about which reasonable people can and do differ (although I would submit that the recent surveys of the empirical literature on vertical integration support the idea that vertical integration benefits consumers in the vast majority of cases supports my position). That said, I don’t see how advocating that regulatory intervention be based on a clear theory and actual data instead of being imposed prophylactically before such problems have occurred makes me a deregulatory shill. In fact, I would argue that it is regulating in advance without systematic data that is the more blatantly political move.

            In short, you have accused me of taking a number of positions that I haven’t taken and have oversimplified my arguments without actually doing me the courtesy of reading my work and engaging my arguments seriously. We have enough real differences of opinion to forego wasting people’s time by misrepresenting each other’s claims and forcing each other to explain what we actually said. I realize that parodying my work and pretending that I have said things that I never actually said may serve your purposes by falsely setting me up as a straw man. But a person who was associated CITP for as long as you were should have more integrity than that. I trust the readers of this blog will recognize our exchange for what it is.

  2. avatar Christopher Yoo says:

    I realized I accidentally deleted a sentence that rendered my first paragraph unintelligible. The first paragraph should read:

    As I have emphasized in my published work, broadband Internet access providers operate in two distinct markets, each with a different geographic scope. The market in which they provide service to end users is clearly local, and this market remains quite concentrated, although not as concentrated as people think. Most importantly here, the merger would neither increase nor decrease the level of concentration in this market at all, as Comcast’s service area does not overlap with Time Warner Cable’s. Consumers would have the same number of options in broadband Internet access providers the day after the merger that they did the day before.

    • avatar Jean Camp says:

      Concentration in edge most certainly increases negotiation power. Peering agreements lack transparency. So while we cannot say in detail how these peering agreements vary or measure the significance of increased negotiating power, simply asserting that this is not the case is pure unsupported (extremely optimistic) opinion.

      In terms of the edge, saying the market is as concentrated as “people think it is” is too vague for a significant response. In fact, I have exactly one broadband provider, Comcast. AT&T DSL does not work here, I tried it. Our local ILEC, Smithfield, does not serve my neighborhood. Local market competition has not come to fruition the way many early advocates, including myself, hoped it would. So while you would map me as having three providers, I have one. If Comcast decided that NetFlix should not stream over Smithfield in order to steam over Comcast, there is no way Smithfield could effectively respond.

      For a very long time I was a strong optimist about facilities-based competition. I thought the FCC was right in initial forbearance. But time and reality have not supported that optimism. I have accepted the reality that it is has not happened. I have changed my opinion because the facts failed to support it.

  3. Rhetorical question: When one sees the phrase “All of this represents well established antitrust principles……” one might wonder whose principles? Brandeis and Douglas principles? Or Bork? Pick whatever side you want….but for anyone to declare something to be well established we have to disregard one school of thought. Well established…for the moment, yes. But not LONG well established.

  4. avatar Harry Johnston says:

    Steve: I don’t understand the basis for the sentence “The FCC is reportedly considering new rules that would permit broadband Internet providers to discriminate against specific content”.

    From what I’ve read about the proposed new regulations, it seems to me that they would allow ISPs to discriminate *in favour of* particular providers (i.e., those that pay them) but not *against* particular providers – everybody would get at least the same level of service as the bulk of the internet.

    Can you clarify this? Is there source material I’m missing?

    • avatar Steve Schultze says:

      Discriminating in favor of one provider is equivalent to discriminating against the others.

      In order to try to claim that this is the case, you’d have to define what counted as some kind of baseline level of service. You phrase this as, “at least the same level of service as the bulk of the internet.” Is that the right measure? Who is to say? How do you determine what “the bulk of the internet” is when a large percentage of internet traffic reportedly comes from those now-favored providers?

      It’s an illogical rabbit hole. Discrimination in favor of one is discrimination against another.

      • avatar Harry Johnston says:

        The distinction between “discriminating in favour of” and “discriminating against” may be fuzzy in some circumstances, but that doesn’t mean it doesn’t exist. (Besides, there should already be a standard for a baseline level of service, regardless of whether or not you have strong net neutrality – particularly in a monopoly or near-monopoly situation.)

        It seems to me the idea of discrimination in a network is inherently fuzzy to begin with, for example:

        If one provider rents rack space inside an ISPs network (or rents capacity from a third party, such as Akamai, with rack space inside an ISPs network) is the ISP discriminating against every other provider? If not, what makes having server(s) inside an ISPs network any different than, say, having dedicated links between your servers and one or more points inside the ISPs network?

        If an ISP purchases extra capacity from a particular backbone provider, or adds a new peering point, is that discriminating against everyone whose traffic goes via a different backbone?

        HOWEVER, since posting my previous comment, I’ve heard that the new rules would allow an ISP to set arbitrary prices for “premium access” – e.g., charge Hulu 1c/GB for “premium” traffic but demand $10/GB from Netflix. (There’s a “commercially reasonable” clause, but I gather that’s too weak to be helpful in realistic situations; sticking it to the competition *is* commercially reasonable, after all.)

        If premium access is necessary for a certain class of provider, which seems likely, then charging differing rates *is* discrimination, and there’s nothing fuzzy about it.

  5. avatar Mitch Golden says:

    I came a bit late to this party, but amusingly I came here just after reading a very nice article on vox.com that explains quite nicely why Mr. Yoo’s arguments are so unconvincing.

    http://www.vox.com/2014/5/6/5678080/voxsplaining-telecom

    But even without such careful analysys, it’s pretty clear why we don’t want to allow the TWC-Comcast merger. There amount of competition in the last mile is between slim and none, everywhere in the US. Clearly, no one is finding it worthwhile to enter into existing markets. (As a good example, in my part of the world, Manhattan, you can’t even get hooked up to Verizon FIOS unless it is already deployed in your building. Verizon has essentially given up.)

    How would allowing fewer last-mile providers improve this situation? If anyone has the wherewithal to challenge Comcast, anywhere, it is a company like TWC.

    I daresay that the management of Comcast knows quite well that the point of such a merger is to increase their market power. If they just wanted to buy TWC’s profits, there are any number of companies.

    • That last paragraph is dead on. Comcast (and the TWC executives who will also benefit) wouldn’t be doing this if they didn’t think it would increase their rents.

      But I think the second-to-last paragraph is mistaken. TWC might have the wherewithal to challenge Comcast, but it makes no sense for them to do so. Far more profitable to have a gentlemen’s agreement about territories and pricing than to get into some kind of ruinous competition. Oligopoly is pretty much the name of the game everywhere.

      There’s one more elephant in the room, though: Comcast isn’t using its market power and local monopoly to maximize revenues from Netflix and other competing high-data-consumption streaming services; it’s using that power to give an advantage to high-data-consumption services that it owns or controls. The company is perfectly willing to give up a few million marginal customers to get approval for this deal, but they’d scream bloody murder if asked to divest NBC/Universal or hive off the cable-distribution side.

    • avatar NathanT says:

      I came to this interesting debate a bit late too… boy where did the month go.

      I guess the disclaimer goes something like “I am not a lawyer,” and I really don’t get the whole antitrust principals. In high school I was taught antitrust laws were in place to keep companies from becoming monopolies where they controlled everything including prices.

      And yet, then I learned that hey cable companies were granted special immunization from all antitrust laws and they were allowed to be the monopoly in an area. Now this was back in the 60s or some time, so long before the Internet at we know it.

      Now, I don’t understand the arguments regarding fringe services, the relationship between backbone providers, data transfers, termination of traffic, etc. etc. So my mind, is all about the consumer.

      And this is what I see. I live in a city where THE ONLY broadband provider that will service my house is Comcast. There is no other option; even DSL won’t go above 128Kb [hardly broadband anymore] in my neighborhood. So, I have to pay nearly $100 for a wimpy 12Mb connection down (3Mb up) and that is a business connection so that I am even allowed to do any upstream content of my own.

      And yet, just a few miles south, in another city, where Comcast is not the only option, but there are over a dozen broadband providers, residential users can get 500Mb to 1Gb connections for less money than I pay; and businesses get 1Gb connections up and down for only a little more.

      That tells me that as a consumer, anywhere there is only one or two options, the consumer is hurt, lower speeds, higher prices, not allowed to transmit as much data…. regardless of all the agreements between interconnectedness. Creating huge conglomerates of cable companies that already have monopoly in local markets, only gives them the power to squash any up and coming competition nationally. Yes, the merger will have a negative impact on the actual consumers (regardless of the impact it may or may not have on other providers such as Netflix).