FTC NN5

From Wiki
Jump to: navigation, search

Discrimination, Blockage, and Vertical Integration Panel

This panel addresses topics such as: The incentives of ISPs to discriminate against or block unaffiliated content or applications; the risks and benefits from such discrimination: and the risks and benefits of vertical integration by ISPs into content and applications.

>>MICHAEL SALINGER
Thank you, Commissioner. As Commissioner Liebowitz mentioned, I'm Michael Salinger. I'm the Director of the Bureau of Economics at the FTC. This will panel we're going to talk about what is, in some ways a new issue with respect to net neutrality but in other ways an issue that's come up whenever we've had to deal with the delivery of content. There was delivery and content. (laughter) Going back to movie theaters at least. There is probably something going back to ancient Rome. We have a very distinguished panel today to help us with these issues. We're going to start out with Joe Farrell. More years ago than either of us would like to admit, I can tell you that Joe Farrell was a great ultimate Frisbee player, which, which was the first time I realized that he understood the delivery and reception of things through the air. (laughter) That is, of course, not the reason we have him here today. He's Professor of Economics at Berkeley. He's also had a distinguished career in government, having served as the Chief Economist both at the Federal Communications Commission and at the Department of Justice, so he's certainly well suited to help us think through the relative roles of regulation and antitrust. Joe? Will you lead us off?

>>JOSEPH FARRELL
I'm going to slide somewhere in here. All right. I'm going to try to go fast because Michael didn't admit to it, but they have us on a tight schedule here. The overview of my little talk, first of all, I think there are real reasons for concern. I'm going to try to be the first panelist of the day. Maybe the only one of the two days who is bottom line on this you will have trouble guessing until the last minute. I think there are some real reasons for concern and opponents of net neutrality regulation who claim that there are no problems are mistaken. I see three real reasons for concern. Number one, for reasons that I'm going to try to explain extremely briefly, and that may well not work, charges by last mile providers to content providers, may in their true economic incidence actually be paid in substantial part by customers of broadband competitors. And that raises potentially some serious competition policy concerns. Secondly, there is a concern, if you allow last mile providers to make charges on content providers, there is a concern about possible ex-appropriation of successful content providers and third there is a possible concern about inefficient or harmful leverage. However, although these concerns are all, I think, substantial and worth worrying about, at least some of them are quite uncertain. The economic conditions for them to be significant problems are not only real conditions, that is, they might not be real problems, but also very hard to observe and pin down. So the real issue is what to do in a case where there are potentially serious problems, but things may, in fact, be okay. And following on commissioner Leibovitz's suggestion of a third way, I'm going to ask whether there is an appealing middle ground. So the first issue that I want to raise is actually not, I think, the most important, but is perhaps the clearest in terms of the incentives. In traditional telephony, we have what's been known as the terminating access problem. That is the following issue. When a phone company charges other callers, or the telephone company of callers, to call its customers, that's called terminating access charges, the question is who pays that? For reasons that I don't have time to get into, with certain common industry practices, in this case, mandated by law, it's not just calls involving this company's customer, whose prices go up, it's all long distance calls. I think it's very possible, although not guaranteed, similarly, if Comcast starts to charge Amazon when a Comcast broadband customer deals with Amazon, that the charges will be borne not just by Amazon or by Comcast customers but also by, let's say, AT&T customers. And this ability to collect money from and or impose costs on customers of your rival, strikes me as, if it's large and significant, which we haven't, of course, established, potentially a very serious concern. The second concern, ex-appropriation. Google is very successful. Will Comcast charge Google to access Comcast customers, in a way who's price is based on Google's' success, and therefore, very seriously risks expropriating some of the fruit of Google's success, charging more for such access since Google has perhaps a higher willingness to pay. Or is it the other way around. Maybe Google will charge Comcast. We don't know. I'm going to skip over the second bullet. Third, leverage. Okay. If a broadband provider is integrated into profitable content and in particular, both of the main typical broadband providers that we have these days are integrated into at the margin profitable content, TV and phone businesses, a broadband provider is likely to resist substitutes, unless it can charge them, unless it can and does charge them, a comparable contribution. Madison River arguably was a case of this. There is a question of, in what circumstance, this is actually inefficient or harmful to consumers. In one circumstance, it merely preserves a historic and not necessarily very appealing pricing model but doesn't necessarily do a lot of harm in itself and perhaps we'll come back to that. So I wrote with Phil Weiser a paper on the internalization of complementary efficiencies, anacronized to ICE, arguing and then qualifying the argument that a broadband provider wants customers to value its product, and at the grandest level that desire encourages good platform management which means among other things, encouraging attractive applications providers. But, of course, that's not the only thing that it wants. In particular, given that it does face some competition, there may well be an incentive for a large broadband provider to weaken independent content providers or sign them to exclusives, in such a way that a smaller rival has less attractive content available to it. It might be worth doing that even if, as a collateral cost, the content would be available to your own subscribers, is not quite as appealing as it might otherwise be. A second reason for concern is the desire which I think will be very strong in this business, with substantial short run market power, to say the least, and very large fixed costs. The second concern is a desire to do price discrimination. Price discrimination as you've probably all heard, many economists say in forums like this, it's not necessarily harmful. And that's correct, given the other alternatives available, but the desire to be able to engage in price discrimination is an important motivator for extending control beyond what is efficient. So how do we think about that trade-off? Okay. So here's the uncertainty predicting what behavior will be like if it's not controlled by some kind of rules, it's pretty hard. Does that imply, as some might suggest, that we should regulate because there's risk of something going wrong if we don't? I think that's the Lawrence Lessig view, things are working very well, let's try to make sure that they don't get broken by misguided selfishness. Or does it imply we should not regulate because things might be fine without regulation and or regulation might provide some problems? That debate tends to be conducted at a very ideological and not analytical level that. That may, in fact be the central debate here. So it would be nice if we could raise the level of that. What should at this time depend on? It should depend on the probabilities of there being a problem but we don't know those probabilities. And it should depend on the ability, if you don't regulate now, to address problems later, or if you do regulate now and see that the regulation is counterproductive, to address those problems later. Can you do that? It's often been suggested or certainly sometimes suggested, that because these problems are in a broad sense competition problems, you could address them ex-post with antitrust. I'll hope to say more about this later in the discussion. I'm not convinced that antitrust as currently enforced is going to do a good job on those potential problems. How do you poise yourself to act, then, if that's the way to go? Perhaps you need to establish some clear understanding of what the principles are, and set up some agency, which could be courts or some other agency, with a will and abilities act. One interesting point here, there are two models of doing that. One is, you're going to prevent problems as they come up, ex-post. Cure them. For that, you need rapid and predictable enforcement. Another is, you're not going to try to do that. You're going to deter misbehavior through some kind of, to put it crudely, punishment strategy. For that you don't need rapid or predictable enforcement, you just need very hard- nosed enforcement. Is there a middle ground? Broadband providers mostly say they want to be able to control harmful content. They want to be able to charge for congestion, and higher speed and so on. Net neutrality advocates mostly say they are concerned about ex-appropriation and about leverage. Haven't heard many of them talking about the terminating access problem but you could add that. There is a gap in there. Does that gap suggest the possibility of win-win rules? If so, an important thing to do would be to explore those, explore how they work and make sure they are win-win-win, where the third win is perhaps the most important, consumers rather than just the participants.

>>MICHAEL SALINGER
Thank you, Joe. We'll turn next to Greg Rosston. Greg is currently the Deputy Director of the Stanford Institute for Economic Policy Research at Stanford. He also had a career at the FCC, where he was the Deputy Chief Economist. He's written extensively about the application of economics to telecommunications, and I'm sure he'll tell us about those principles now.

>>GREGORY ROSSTON
Thank you. It's sort of tough going after Joe because I agreed with almost everything he said. What I'll do is I'm going to expand on a couple of ideas that I had that will hopefully complement what he's said on these things. And express a little bit more, as a guy from silicon valley, I don't have any PowerPoint slides. (laughter) So first thing, for me, and I think an important question for the FTC, what is network neutrality and what does it mean? You gets lots of different definitions. That's a key question in order to say should we do something? What is it are you planning to do and what do you mean? I think that was -- that's part of what, what Joe said was, should we do something? It depend on what do something means. In addition to the probability of there being a harm, what's the option you're going to take in order to do that? This debate has sort of been, you know, about, again, you've heard the horribles on one side versus the horribles on the other side. Both of them -- both side overstate things when I hear that cable companies and telephone companies say, there is no need to regulate because we don't do anything bad and we're not doing these things you're going to regulate and prevent us from doing, then you shouldn't have a problem if you say you're not going to do them. On the other hand, you know, there is this problem of regulating and not knowing what the incentive effects are from the regulation. You need to think about both sides of this and try again. I think that putting this as a more dispassionate argument, thinking about what people's incentives are and what are the effects of regulation, are really an important way of doing this. To complement Joe and Phil Weiser, I recommend that everybody read their ICE paper. It really sets forth the incentives for vertically integrated firm or vertical restrictions by firms so you should definitely read that paper if you're interested at all in this issue and the economics behind that. IN that one of the things in traditional antitrust economics has been that for vertical integration, vertical problems, vertical issues to arise, you've generally got market power at one level in this. We have lots of pro-competitive, vertical relationships even when there is market power at one level. But when you try -- when you get away from market power at one level you tend to have less problems in vertical relationships. If you don't have market power. The one caveat to that is the point that Joe brought up about the terminating monopoly. In some sense you do have market power still in this terminating monopoly but in most other instances, this vertical relationships issue goes away substantially if you have competition instead of market power at the level. But when you do have market power you do have issues and I won't go through them, these vertical relationship problems, incentives for somebody to exploit it, they can expropriate and cause problems for innovation. In this case, I think the key that I want to focus on is that, one of the key policies that should be promoted is how do you get rid of this duopoly that we have and yet -- triopoly -- quadopoly -- whatever you want to call it, and get more competition to at the level of bringing broadband access to peoples' homes, and that would be a good way to solving a lot of problems. This have been people who are concerned with this terminating access problem. I want, you know, one of the ideas is that you look at Europe and the cell phones, how extremely high terminating payments for calls. But one of the things that happens there and this is write slightly disagree with Joe, if you call a land line phone in Europe you pay different price than if you call a mobile phone so there is a price difference in these terminating access problems that we have in rural telephone companies in the United States. There was no price difference. So that way you leveraged it onto other customer who weren't actually calling those people. So if we get -- generally, though, getting more spectrum out, trying to reduce restrictions on broadband over power lines, but the key is making sure, for example, when we get more spectrum out that we actually enforce the antitrust laws and make sure that we have the ability to have multiple competitors providing broadband access to the home, and that's going to help alleviate these concerns. In my mind, this is a much better way than trying to mandate network neutrality. One of the other things you want to think about is, if you're trying to encourage competition, and encourage new entrants to come in you might want to let them -- you probably would want them do as much as possible in order to have the returns to their investment. And some of these vertical relationships that people are concerned about, that may increase the profits of a new entrant, may be the thing that's necessary in order to get a new entrant, in order to compete. You may want to think about how do we balance between regulations on incumbents versus new entrants. There may be a justification for differential regulation, if you think there is a problem and your whole goal is to encourage new entry. But on the other hand, you do need to have incentives for the incumbents to upgrade their networks as well and to try to provide higher speed access. You want to make sure people have incentives to upgrade but also not to have incentives to take advantage of customers and to forestall innovation. I realize this has been a high-level talk and not coming down on a particular side but trying to highlight the issues that you need to be concerned with in thinking about what regulations might be. The other thing that I want to talk about is, what do you do as a regulator? What would one do if you said, you need to think about whether you want to institute a regulation, ex-post or ex-ante. Ex-ante regulation or ex-post enforcement. You want to think about the two pieces that Joe said, the probability and the efficacy of ex-post enforcement, the probability of harm but you also need to think about the relative effectiveness of Ex ante regulation and also ex post enforcement and also what incentives these create for firms to provide services to consumers. So my bottom line is, try to increase the competition to get rid of the market power problem at one level and worry -- think about the terminating access problem but I'm not sure that's as big a problem as making sure that you get competition there. I think I'll leave it at that.

>>MICHAEL SALINGER
Great. Thank you. Our next speaker is Simon Wilkie. He is the Director of the Center for Communications Law and Policy at the USC Law School, and Professor of Economics at the Enbark(ph) School of Communications and also a former Chief Economist at the Federal Communications Commission. Thanks.

>>SIMON WILKIE
I'm going to do a multiple media presentation to myself. I've got slides, I've got notes, and I have props. And what I want to do is, three things in this talk. The first thing is, I want to follow up on Joe and Greg. Basically I'll be singing the same tune, so y'all heard it before. In fact, Commissioner Liebowitz set the tone exactly right. Let's find the third way. Then one of the requests that was put to us is, if there is an issue what do we look at? What data should we be looking at? I'm going to give some examples of what the data looks like and where you can find it. Where the bodies are buried, if you will. And third, I'm going to discuss policy alternatives in light of what the current real world situation is. There are frankly, a lot of nonsense written on this topic, particularly in Washington, D.C. Because I only have 10 minutes. I'm actually going to start with the punch line and work backwards. I think the punch line is, as everybody has suggested, the rhetoric in D.C. on both sides is to the extreme. It's not really reflective of realities so therefore one extreme I think I would not support, for example, the Marquis (ph) bill. I think the Marquis (ph) bill causes a lot of harm, potential harm by discouraging potential innovation. It's also fairly badly worded in terms of not defining terms in a transparent manner I feel. I'm not a lawyer, I'm just pretending to be one. On the other hand, I don't accept the proposal that this is a competitive market and it can be completely deregulated. In particular, for the issues that Joe and Greg have mentioned, that the terminating monopoly is a real issue. It's an issue in every telecommunications markets where people interconnect. The key point is, in the formal language of economics, when we study two-sided markets, consumers at one end, single home versus multi home. I'll talk about a subtle difference when I was at the FCC, when I felt we could completely deregulate a market versus why I think that's an issue here. It has a little bit to do with the difference between single homing and multi homing, something that again Joe has written on. I also want to emphasize a point Greg raised, and it's a good point for people to start with, is to read Joe and Phil Weiser's paper. (inaudible) (laughing) I would also suggest that people also learn a lot by looking at the works of Pat DeGrebber (ph) while he was at the FCC where he studied bill and keep and the rationales for interconnection regimes based on bill and Keep and Atkinson and Barnakoff (ph). It turns out, for a surprisingly wide level of situations, what we have today is actually optimal. So that suggests that perhaps we don't have to do much. So that's my punch line. But let me back up to where we are. Before that I want to give a pitch, as was mentioned, I'm at the Edinburgh Center amongst other things. The Edinburgh Center is a fantastic location with lots of money and beautiful buildings in Los Angeles. One of the things that we like to do there is to take people out of D.C., wine them and dine them. We recently celebrated the 10th anniversary of the '96 Act with a dinner and theatre and a couple of cases of 1996 wine. Seems now we're looking backwards we thought we should have French wine. Sorry. So one of things we did is we locked the industry participants, who were normally in Washington, D.C., in this nice environment to see if we could come to some consensus. Remarkably we almost did. The principles, that people -- at one level or not would agree with, not a complete consensus, as I mentioned are our website which is cclp.usc.edu. Basically, the idea is to sort of modify, if you will, Michael Powell's four internet freedoms to say rather than enforcing nondiscrimination that essentially the gist of the proposal, consumers should have the choice of a net neutral package being offered to them. We should establish a floor, a baseline level. For instance, at 1.25 megabytes per seconds we can deliver that pretty much to everybody in America, with the current infrastructure that we have now so we're not discouraging any new potential investment. So if firms offered Vanilla, which is a net neutrality package that everybody can get today and then above that level, all bets are off, that would be one approach. The behavior might be you might want to add the tiering and offering higher levels of prioritization, are allowable but they would have to be offered on a nondiscriminatory basis or what economists call second-degree price discrimination, that is, the prices are functions of the level of functionality offered. Not the identity of the customer. So that would in particular exclude foreclosure. So there was a fair amount of consensus to support those type of ideas. So working backwards from my conclusion to my second point, which is where are the bodies? We mentioned the terminating monopoly problem and Greg alluded to the issues in Europe, the data can be found at the end of the FCC's wireless competition report where it compares countries where, like the United States and Canada, and Hong Kong, where the user pays or I pay for every call coming in, I actually see the entire cost of my phone bill, right? So if my phone company, if T-Mobile decides to raise their rates to me, then I can switch to Verizon. There is a lot of competitive for me to shop around. On the other hand in countries in Europe, the billing goes the other way. That is the person who initiates the call pays the termination fee. That way if my carrier raises its termination rate, it doesn't affect me. I the customer as Joe pointed out have no incentive to turn, so let's just change if you will the property rights of who is paying for the call, not really a regulatory issue. And the equilibrium changes from a competitive equilibrium to a monopoly equilibrium. There is a great paper written by John Jacques LaFont (ph) and Patrick Gray (ph) in the Rand (ph) Journal several years ago which explains how this change in the rule gets you to the monopoly equilibrium. This is an endemic problem in Europe and I think, Chris Libertelli is in the room, if you use Skype, you see Skype to Skype calls are free to Europe. I think to a land line it's three cents. As Greg mentioned, there is a disparity in the termination prices. And the disparity shows up in the Skype pricing. I think it's 3 cents to a landline and 21, these are Euro cents -- oh it's 2 cents -- Christ has two fingers up there. So it's 2 cents to a landline and 23 cents to a cell phone? So the terminating monopoly price is 21 cents. How does this impact consumers? Well, in the U.S., we pay, my data is a year old because these slides are a year old, an average, carriers receive revenue of 8 cents per minute per minute called. It's now down to 7 cents. If there is anybody here in the wireless bureau and the latest competition survey? The U.S. consumer yaks on their phone for an average of 680 minutes per month. Some of the cheaper carriers are even higher. If we go to the UK, which also has five large carriers. It's basically, prices are four times higher and consumers use the phone 1/4 of the time. So we go to an average revenue of 21 cents per minute, from 7 cents a minute, three times higher, and they use the phone for about 150 minutes a month. Germany it's even worse. So basically just that single change of where the fee is recovered for terminating the call, the closer you get to end point, the bigger the problem. We can look at that as it's the same across every country in the world. In the U.S, Greg also mentioned, we have the problem with the rural Irex (ph), and the recip comp order. We had the problem of the Selac (ph) money pump that I could set up a Selac (ph) just take ISPs as my customer, set up very high termination charge. By the way, this is going on at the moment. Does anybody use freeconferencecalls.com? This thing going in space, exactly, do it for free because they are stiffing your carrier upon the termination charge. The other example is in the international settlements arena, one of our conferences, we had the former chief of the international bureau saying how these international competitors negotiated the deals where the U.S. would say, how about a dollar a minute and the other country would say, why not $2. And the U.S. is $2.50. So we bargained our way all the way to the monopoly price. The U.S. became enlightened and became the force of bargaining these calls down to zero, which comes out to be the efficient optimal price in many cases. That led to a new phenomena which was called whip slurring (ph). International arena, the U.S. fights this in the WTO. It's where a country plays one U.S. carrier against each other trying to offer monopoly rights to reach that country. So it's exactly the type of foreclosure that certain opponents of net neutrality say could never happen. We deal with it all the time in telecom. The FCC can look at these cases. My final example that I wanted to talk about is what happens in Australian in the Australian cable market. For instance, Australia was late to the cable market. Australia is very similar to the United States in term of population and demographics. The U.S. we have 88% market penetration of cable TV or satellite for pay TV MVPDs in the arcane lingo of the FCC. Australia decided to go one better, rather than licensing monopolies they would license duopolies so we would have two competitors. However, in the U.S., we have what are called the program access rules, which says a cable company couldn't foreclose its competitors by buying programming vertically integrating and not selling that programming to its competitors. This has been very effective to all the spur competition. It's a regulation, no doubt about it. Australia did not have that so what happens in Australia? You have two cable systems, both with half the channels, if you want to watch both Carey Grant and Humphrey Bogart, you're out of luck. What's the market penetration in Australia? 22%, so again, you get this diminution of consumer surplus by a huge margin. So the argument that you can't have this emerging, this vertical foreclosure emerging in equilibrium is just nonsense because the incumbents have the maximum incentive to differentiate the product. That said, and I'm out of time, on the contra side, there still is this issue of providing enough money to incent the last mile investment. So therefore, try and stop the absolute abuse of monopoly power but don't stop charging a premium for enhancements. Thank you.

>>MICHAEL SALINGER
Thank you, Simon. For people concerned about foreclosure of content by delivery providers I would observe the FTC is the deliverer of this conference and we've had a lot of people from the FCC, former officials from the FCC, but we do have one former FTC official, Tom, briefly, Tom Lenard. He works at the FTC as well as the OMB and the (inaudible) Wage and Price (inaudible). Currently he is a Senior Fellow and Senior Vice President at the Progress and Freedom Foundation. He's written extensively on telecommunications issues including a recent book about net neutrality. Tom?

>>THOMAS LENARD
I think I'm going to be less of a third way type than the previous speakers. If that's the way they describe themselves. However, I will join them in heartily endorsing the Phil Weiser paper. It's a great paper. I think question is really, is whether we ought to be concerned about what really are pretty hypothetical concerns about market failure and market power. Those types of issues, when it's not even clear yet what liable business models for broadband will look like and how we'll be able to develop viable business models, that will cover the really large costs of building out the infrastructure. It looks to me there are three salient facts about the broadband. One is that it is a very young business. If not still in its infancy, not very far out of it. The second is that it's a distribution business. And the third, that it's a business that is characterized by very large upfront costs. So as the industry evolves, it's unclear what that viable business models will look like for this industry, but arrangements that might be viewed as not neutral or discriminatory are very common in the distribution business and they are very common in businesses in which there are a -- a large portion of the costs are up front which is the case with both the broadband distribution business and with the content that it delivers. And, in fact, if such nonneutral business models may very well be essential to provide sufficient revenues to cover the costs of these investments. In addition, some viable business models are almost certainly going to require that broadband be bundled with Comcast, which is very typical of distribution businesses. So what may be needed for successful business models, may be a bundled product offering that is sufficiently attractive, to attract enough consumers to become subscribers at price that will pay off the cost of these very large investments. While these bundled broadband content business models may be needed to drive the necessary increases in subscribership it will also be the case that consumers will demand broad access to the internet and the content that's available. It's very common for vendors in distribution businesses to sell consumer goods and services to consumers, to sell their own product and services along with those of other vendors. Competitors' content can increase subscribership at very low or perhaps even zero in marginal costs so it won't be in the provider's interest to block content that consumers want and thereby lose subscribers that are going to be high margin subscribers. I think it's also critical to think about net neutrality regulation in terms, as mentioned before, obviously, of its effect on entry. So the ability to bundle, make exclusive deals, otherwise have nonneutral business models, may be key to facilitating entry. A possible example is the Clearwire Bell Canada deal in which Clearwire entered into some sort of an exclusive deal with Bell Canada to provide its voice services in exchange for a $100 million investment. Clearwire doesn't block other VOIP providers apparently, but assume for the sake of argument that it does discriminate in favor of its own VOIP provider in some way. A net neutrality requirement would preclude such a deal and might deter a company like Clearwire from entering the market as a new platform to compete with the incumbent, the incumbent platforms and certainly would make, would make such entry more difficult, which is exactly the opposite of what we want to do. Of course, all of this is before we consider capacity constraints because obviously under congested conditions, efficiency is going to require charging positive prices and some of the pricings might also be considered nonneutral or discriminatory from a regulatory perspective. What if there is insufficient competition? In my view, at the present time, even with the relatively small number of competitors there is pretty intense competition for customers and the recent FCC data indicates that that competition is growing pretty rapidly, and the most recent report, 11 million mobile broadband wireless subscribers. But even if broadband was a monopoly, the case is pretty tenuous. Again, I read the Phil Weiser paper. It was one of the first things I read a couple of years ago, when I was starting to think about the net neutrality issue. It's a nice, clear paper and I was convinced after reading the paper, this was not a real problem. Unfortunately, neither of the authors of the paper were similarly convinced. But the ICE, the central core from ICE, which is essentially kind of a follow on to the one monopoly rent theorem, claims "even a monopolist has incentives", this is a direct quote, "to provide access to its platform when it's efficient to do so and deny such access only when access is inefficient". So it's not in a monopolists interest in general to try to monopolize an adjacent market and exclude competitors applications. What about the exceptions, which were more persuasive to the authors of the article than they were to me? The one that steams to be the most relevant is the one where, you know, you have a competitor in the adjacent market which can threaten the primary monopoly. This is what the Microsoft case was all about. The court found Microsoft had undermined the Netscape browser, because of concerns that it threatened Microsoft's position in the operating system market. And similarly net neutrality proponents sometimes argue that broadband providers that are dominant in the video or voice markets might discriminate against independent video or voice override, which could potentially occur but it seems unlikely this will occur when there is at least some competition in the market. It's hard to envision the Microsoft campaign against Netscape, if there had been even one significant operating system competitor. Of course, when you're talking about new entrants, new entrants don't have any primary monopoly to protect so that exception is completely inapplicable and bundling voice or video with broadband may be the only strategy that makes entry fees feasible. This whole debate net neutrality is frequency cast in terms of its effect on innovation and the proponents focus on the harm that compromising the so-called end-to-end principal would cost innovation which they maintain occurs at the edges of the network. There is unfortunately a striking lack of concern about the effect on incentives to invest and innovate in the network itself where broadband providers already, as an indicator, are spending tens of billions of dollars and where the engineers tell us a lot of innovation is already and will be occurring. But the advocates of net neutrality raise the specter that applications and content innovators will be deprived of a way to get their new products to consumers and therefore will be discouraged from innovating. It's difficult to envision this happening in the current broadband environment. First, there is intense competition in local markets. Even sometimes when there are only two providers. And that competition is growing. So a provider who denies access to content or applications that consumers find valuable will reduce the demand for its services. And moreover the market for content is not the local market. It is really a national or even international market, and so it's very difficult -- it's difficult to envision a case where an innovator will not be able to find some outlook for an innovation that's truly worthwhile. Finally, of course, this has not been mentioned much, there is a well-known distortions associated with common carrier regulations which is what net neutrality really is. So it's really, in my view, much better to apply some sort of a case-by-case approach for alleged abuses, to attempt to sort those out that are really, act like they are competitive.

>>MICHAEL SALINGER
Thank you very much. Our final speaker today, literally, wrote the book on regulation, or at least the book on regulation that many of us had to read when we were economics students a long time ago. He had a distinguished career as a regulator, having chaired the New York Public Service Commission, and he had the ultimate regulatory task of being the Chairman of President Carter's Council on Wage and Price Stability but he is perhaps best known as a deregulator for his stint as having been Chairman of an aeronautics board overseeing airline deregulation. He is Professor, I assume Emeritus, at Cornell.

>>MALE SPEAKER
Formerly meritorious.

>>MICHAEL SALINGER
(laughter) Yes. So we're of course very pleased to have with us today, Dr. Fred Kahn.

>>ALFRED KAHN
I have to make, by virtue of necessity, take advantage of my comparative advantage, which is age. The thing that I find most distressing about the movement towards network neutrality, apart from at least until recently, its lack of clarity, oh, I'm sorry. Can you hear me now? That it seems to me to be running the risk of what I've always accused regulators of, which is of having a very high large known propensity to meddle. I must confess, I'm going to bring the wisdom of age, and therefore, a lack of the ability to weigh some of the probabilities that have to be weighed if one wants to take seriously the arguments of the opponents of network neutrality. I've participated in a time in which we had a remarkable convergence of people who believed in competition and huge, really, diversified proponents of getting the government out of the way where ever it seemed remotely possible that the competition would work. Let me say in advance, I did, however, have some exposure to Joe Farrell's thinking I've not ignored my thinking, the danger of terminating a monopoly. My very strong inclination along with Rosston and Tom is to get things right in the first place. That is, to recognize the high degree of ignorance we have about what kind of problems will emerge, if any, number one. Number two, recognize that a good deal of the advocacy of network neutrality is economically ignorant. Certainly, insufficiently cognizant of the kind of consequences of regulating a market that is becoming increasingly competitive. Let me warn you at the outset, I did have enough contact with the problem of terminating monopolies so as to have a suggestion, at least in the interim, until our ignorance is more nearly dissipated, that may handle diverse concerns of the advocates of network neutrality. I certainly begin with a very strong presumption in favor of the deregulation by pointing out, first of all, I'm the only person in this room, I'm certain, whoever took a course at Joseph Schrimpater (ph). (laughter) How's that for a qualification? Particularly in the circumstances that he envisioned, which no one could deny, we have the most extreme example of competition by innovation, the wisdom of being very careful about interfering with those incentives. I know every liberal reform in the last 100 years, and I was a student of the populace of the progressive movement of the turn of the 20th century, understands that every reform that's ever been proposed, met with the objection that it would interfere with investment incentives. But I think also that it's clear that this particular industry, this dynamic kind of competition, is certainly as close to unique as any could be. So I think the lesson of history, be very, very careful that you don't meddle with a process that's clearly characterized by sufficiency of competition. Now, of course, there is no certainty about the sufficiency of competition or the duopoly. By the same token, it's possible to observe the presence of competition. And I testified for telos in Canada that said, well, we'll talk about deregulation only where we see competition. Particularly facilities-based competitions, because facilities-based means low marginal costs. Therefore, competition once begun is not going to be quickly abandoned, particularly when one entrant has a very small part, and has the facilities in place, and in those areas in which the facilities can be reached and where we have competitive behavior, there above all else, the wisdom, I think, of experience is to wait and see what kinds of problems emerge. Now, the only -- case I know that's been cited as an argument for some sort of regulatory intervention, is the Madison River case. And a more obvious case of the abuse of a vertical position, I cannot imagine, of course, it was properly treated. Peremptorily, both in the United States and Canada. That does not get to the sufficiency of competition in other situations and the question of the definition of exclusionary tactics. But this, the one that everybody cites, is the most obvious case for which there is the most obvious answer, I think. A second observation I want to make is that so far, in my reading of the literature on network neutrality, I don't see any perception of the meaning of discrimination. The question of what is discrimination and what is not discrimination. And the opposition to tiering, which seems to be at the heart of the proposal. I have never seen anybody answer the question whether tiering is discriminatory. We know certain uses require much more instantaneous connection. Forgive me for using old descriptions than voice, VOIP. We know that that is particularly demanding. Just as, God, I hate -- now, I can't remember what the hell I was going to say -- (laughter) I don't see that there is any -- anyone has confronted the question of whether they are therefore charging for this preferential delivery is really discriminatory or whether it does not involve in the short-term opportunity costs, they are slight, I gather, but it does mean that the people who use email, will have slightly less rapid delivery, but without necessarily interfering, there are only certain uses that are particularly demanding of the immediacy of delivery. Take the logical case of remote medical analysis prescription and treatment. If it had opportunity costs in the short run it's not discriminatory, to charge more for it. If it has long-term costs in terms of necessitating more investment in broadband, just as, let's say, video could accomplish by telephone companies with cable companies requires more broadband, then again, it's not discriminatory, if it has higher short-term opportunity costs and long-term investment cost. The one aspect of the network neutrality case that does seem to be demanding of attention is the one that you've described as the terminating access problem. I had certainly until fairly recently thought that the presence of competition among originators access to the internet would be sufficient to protect a Google in the assurance of access to ultimate customers, and it was only in time that I became aware that while I saw every reason to charge originators of content in contrast to some consumer advocates who shall remain nameless, no, all the favorite should be made by the ultimate subscribers, the analogy to me, if I can think with analogies, would be newspapers. Would you say that newspapers should be prohibited from charging advertisers? And should get their money entirely from the people who buy the newspapers? Well, there is the two-sided market, and it's obviously absurd to say that. The same thing is true of Google. The originators of content, who want access to the public for purposes of advertising. So again, it's absurd to say charge only you and me, who are the subscribers to DSL or cable modem service. So the two-sided charging is necessary. That does, I thought, the presence of competition at the originating level, I thought, would be sufficient to protect content suppliers, and the enjoyment of the fruits of their innovation, which I think, of course, again, is very important, and I must say this reading some of the two sided market literature and then just looking at Joe's absolutely inscrutable comments, I saw that I was assuming, I was making a factual assumption that was incorrect, which was that the protection of competition at the originating level was sufficient to protect Google from exploitations, to use a really inappropriate word. Not realizing that if Google had sent us down one, whether it's AT&T or Comcast, vowed one to put it on the internet, that was sufficient to give it the access to the market that it required. Fortunately, I have a grandson in law who understands it much better than me, and said no, you have to have some sort of what they call peering agreement which is that in order for Google to have access to all possible users of its services, if it originates with one, that does not assure -- if it originates AT&T that does not ensure that it gets carried by Comcast and reverse. The light came on over my head and said, well, what you're really talking about is mandatory interconnection. That, I gather, the word is peering. I thought peering was secretly looking in places. (laughter) Looking. And so I see, again, I think partly influenced by Joe, that mandatory interconnection seems to be the necessary element to give Google the protection of competition that it requires, and give it the access to the market that, in some sense, it deserves, and avoids being held up at the terminating end. Of course, that fits with my basic argument that antitrust can and must be sufficient to handle, and remember, mandatory interconnections is an old antitrust doctrine. As far as I can see, antitrust, I hate to take the responsibility away from the regulatory agency, just as I don't like the Trinco (ph) decision which leaves it entirely with the ex-post type of damage approach. So I see the intervention of the CRTC in the Madison River kind of thing and the FCC as very important, because it can be done expeditiously. Finally, of course, if you read the latest article, you will see a repetition of my ancient argument in a book I published with Joel Duram (ph) 53 years ago, called "Fair competition," the law of economics and antitrust policy. That fairness of equality of competitive opportunity is the most important aspect of antitrust, and clearly has to be applied in this kind of a situation, in the most preemptory way possible, and that, of course, means applying the efficient component pricing, which I call the principle of competitive parody in the cases of people competing with a vertically integrated firm of which they are also dependent. Beyond that I'd say for God's sake, don't tinker. We're moving into an age in which liberalism is becoming converted into what people call progressism. You're going to find that the progressives have a very high marginal propensity to meddle. I'm a 18 th century -- 20th century liberal and it was a combination of those two that played such an important role in deregulation movement. And of that, I'm still unregenerate. (Laughter) (Applause)

>>MICHAEL SALINGER
Thank you. One of the themes that's come up is the relative role of regulation, Ex ante versus using it antitrust ex-post if problems arise. Joe, you expressed some skepticism about antitrust, Greg threw in a little jive about, well, if we enforce the antitrust rules properly. Joe, why don't you tell us a little bit about your reservations about whether antitrust is up to the task.

>>JOSESPH FARRELL
Yeah. I don't think antitrust would even take a whack at the terminating access problem. So I think we're dealing with the ex-appropriation or leverage problems. I think if you tried to bring an antitrust case these days, where you say, we were successful. This firm that we essentially have no option but to deal with, is charging us a lot because we're successful, and that in the long run is going to weaken our incentives to innovate and be successful, I don't think would you get past summary judgment. I think the opposing lawyers would say we don't have to deal with you at all and so go away. So if there are doctrines of fair competition, and maybe this is an FTC thing more than a Justice Department thing, maybe section 5 of the FTC Act could and would step in here, I'm not enough of lawyer to know, but I think the Sherman Act and other primary antitrust statutes are not going to do much. I think, Fred, you said Madison River was clear-cut but you also said ECPR. So I'm wondering, if Madison River had said, yes, you can use Vonage, but you have to pay us our quasi profit for each minute of use of voice telephony hat you don't use because you're using Vonage, would that have been okay or not okay if your thinking?

>>ALFRED KAHN
It sounds to me as though it would not be okay. I know that the Vonage decision, the intervention by the FCC didn't handle the terminating monopoly question. But what I need persuasion, in full recognition of the dangers of regulatory meddling, in this situation, whether the imposition of a requirement of interconnection in effect, would not suffice to handle the danger of exploitation of -- at the terminating end. But, of course, that's on the assumption that the competition at the initiating end is sufficient. I certainly can't contend that -- I assure you that the competition is sufficient but it's such a dynamic situation, that I think the cost of trying to impose regulations at that level, I don't think -- here's an area of which I'm totally unqualified. So let me qualify what I'm saying, but if it's true, that in a very short time, we're going to see the probability of broadband over power lines, and we already have hundreds of cities that have hot spots, WiFi hot spots and Sprint Nextel is talking about spending, in conjunction with who is it -- spending several billions -- $3 billion over the next 10 years, extending a nationwide WiMAX facility, for Christ sakes, keep out of it. Expropriate them later if you will. Don't do it in advance.

>>MICHAEL SALINGER
Greg, you said something about making sure we enforce the antitrust laws correctly. What's essential that we need to take the hands-off now and rely on antitrust later?

>>GREGORY ROSSTON
What I said was make sure we enforce the antitrust laws was in response to making sure we get entry in wireless and it's not that new companies, we have multiple companies providing wireless access, it's not that we have the same companies providing both wired and wireless access. That is what I was saying, making sure we get additional competition at the last line -- that was the point about the antitrust laws being enforced correctly, that I was trying to make.

>>MICHAEL SALINGER
Tom, Simon talked about the Australian experience with cable, why -- how can we be so sure that we're not going to have similar problems with the internet?

>>THOMAS LENARD
Well, I guess what strikes me, again, about this, is we're talking about kind of -- we're talking about hypothetical problems or things that may have happened elsewhere. No one has really mentioned any problems here other than hypothetical problems. Things that could happen, you know. They are theoretically possible, that they might happen. But, it seems to me, to, in the real world, the way regulatory agencies work, with all the pressures on them, all the rent seeking and all the imperfections in them, to start to institute a regulatory regime that nobody has a very good idea of how it would actually work, to solve problems that are hypothetical, it seems to me to be just, I just don't see any way that that could turn out, you know, happily.

>>MALE SPEAKER
I realize --

>>MALE SPEAKER
I think the wait and see approach is likely to be a good approach if conditional waiting and seeing actual problems, we are prepared in advance and know how to some extent, to jump in and deal with them effectively then. And I'm not at all convinced that antitrust in its present state would deal with the concerns that are expressed by those who are concerned here. Is there a way to do a wait and see model that would work well using some other set of principles? I think that would be worth exploring.

>>MALE SPEAKER
Antitrust is literally an imperfect world. Do we know any better how to solve the hypothetical problems than that? If we don't, wait and see.

>>MALE SPEAKER
I don't think people have been talking about it because I think people have been saying let's wait and let antitrust do it.

>>MALE SPEAKER
Sort of give a little bit more about, cable television started out by investing in programmers and they wanted to get content on the cable systems. That was one way of sort of, vertical relationships between the cable programmers and the stuff that (inaudible). Now we have a very different situation in cable. One of the things, some people may be concerned or not about something like the NFL Sunday ticket, which is an exclusive deal with one provider of broadband services, and it may be the case, you think, NFL decided that this is what they wanted to do. It could be in other cases you might have somebody, if AT&T or Comcast says, no, you can only be exclusive on us. So if you were concerned about these vertical relationships these are examples that may come up that people might be concerned with, about how vertical relationships might work. Conversely, you think about the AT&T case and what happened in the AT&T case. It was about having a separate choice of long distance provider. Right now, almost no one in this room has a separate long distance provider for their cell phone? It's a very different thing. Things have changed a lot, and the integration of long distance with your cell phone doesn't cause a vertical problem in most people's mind because you have a choice of five providers.

>>MICHAEL SALINGER
Simon, you were trying to jump in there.

>>SIMON WILKIE
Oh, yeah. Again, I disagree with the description at least in the hypothetical because again we're talking about one level of network connections, so we have to look at what history tells us. However, I would emphasize that we should proceed with caution and pretty much I agree with the policy recommendations of Fred, which is that smart economists know exactly what the structure of the problem is and we've dealt with it before. Fred dealt with it in the airline industry. If you have competition in take-off, but a monopoly in landing, not so good. (laughter) That's why Fred got it right. You don't want to terminate a monopoly on the landing end. But to go back to Joe's point about ex-appropriation, we also have evidence right in front of us today, which I'm going to use my cheap economic stunt, if I go into the streaming video business and want to show goofy videos to people online, if they want to access them with this device, it's perfectly fine. Why? Because of the legacy Title 2 regulation, much of which would be swept away, if I want consumers to have access this content with this device, which is under a different regulatory framework, then the carriers in the U.S. demand 50% of my revenue to get on to the device. Suppose I invent a product with a 40% profit margin that creates huge benefits to consumers. It's not going to be delivered to this device in the equilibrium of the United States at the moment. If we go to Korea, I just had an international conference of regulators at my center. In Korea, the number is 20%. In Japan, number is 11%. Now, this is actually where I agree, I would say, is there cause for regulatory action and intervention and should we start regulating content on these devices? I would say no. The problem is the line spectrum. It is an identifiable problem. It impedes innovation and we know it by looking at it in the difference of the video content available for this versus the video content available for this. So it is a real problem. Don't deny that it's hypothetical. I don't think it actually requires regulatory action so in that sense I agree with Fred.

>>MICHAEL SALINGER
A question from the audience.

>>MALE SPEAKER
Greg --

>>MALE SPEAKER
Fred, you said all we need to do is mandate interconnection. But that begs the question, interconnection at what price? If we mandate interconnection, are we necessarily going to get into messy price regulation?

>>ALFRED KAHN
I don't know the answer, but under the peering arrangement, do you know how the pricing, the charging by the connecting carrier is arranged or is simply this kind of bill and keep.

>>MALE SPEAKER
It's bill and keep at the tier one level. The big eight guys who run the backbone do the linking.

>>MALE SPEAKER
They agree to it themselves, it's not mandated?

>>MALE SPEAKER
It's not regulated, correct.

>>MALE SPEAKER
Not regulated.

>>MICHAEL SALINGER
This is a question for anyone. Do lock-in or substitutability matter? Is it better to reduce lock-in or increase substitutability from government action rather than to regulate behavior? Greg, you were talking about encouraging entry. Maybe I'll go to you on that.

>>GREGORY ROSSTON
When this debate first started as open access in this debate, actually, the debate started in 1887 with the ICC Act. Common carriage. But in this, people talk about you're locked into your cable carrier of your DSL provider. I don't think the lock-in effects are that high. I've switched between cable and DSL. A lot of people have switched. Lock-in doesn't seem like a big problem right now. It seems like Google has solved the email address for lots of people. They are not locked into their email. It doesn't seem like a big problem and shouldn't be a big issue in this debate.

>>MALE SPEAKER
Anyone else want to pick up on that?

>>MALE SPEAKER
Joe is tempted.

>>MALE SPEAKER
No current wireless connection including Q satellite network offers internet connectivity. So --

>>MALE SPEAKER
That's accurate.

>>MALE SPEAKER
Read the card.

>>AUDIENCE
Read the card.

>>MALE SPEAKER
They all restrict traffic types that are allowed. How do we get true high-speed internet connections when providers restrict both in wire and wireless markets? Terms of service for most wired restrict port access just like wireless now. I don't think that was the question. So I think we'll move on.

>>MICHAEL SALINGER
This is for Joe and then for the panel to comment. Do you agree that a goal of our policy should be to encourage investment by the network providers? If so, do you also agree to be successful, selling prioritization service depends on the perception of the content provider at the level of service it will get if it does not by prioritization. That would be interior and inadequate? In other words, does allowing network providers to charge for priorization create an incentive for them not to invest in their networks in order to earn more for prioritization?

>>JOSEPH FARRELL
Yes, I certainly this it's a goal to encourage investment by network providers. A couple of members of the panel have commented on that. I would just say one thing about that. I don't disagree with that. I agree with it. But one of the things, as I think Gigi said this morning, one of the things that's been rather special about the internet is that we really have seen a dramatic success of the openness and opportunity model, which one can, to some extent, contrast against the control and incentives model. So on the internet, there has been a vast amount of innovation that an economist would look at what's going on and say, those people have very little incentive to write for Wikipedia or set up an interesting blog and yet they are doing it. One of the lessons of the internet has been -- a lot of people actually enjoy creativity, and although as an economist, I certainly agree that there are kinds of innovation for which you really do need to make sure that the financial incentives are there, I also think it's important to remember that openness to many, many millions of people doing little stuff off their own bat is quite important. Now, I think the question on the card was, to some extent, about whether price discrimination creates an incentive to wantonly or at least irresponsibly not invest in the low-level capacity, so as to be able to charge extra for the higher-level capacity. There are two conflicting economic forces here. On the one hand you want the product to be good for everybody so that you can charge everybody a lot. On the other hand, when you're doing what I call price discrimination, I know Greg Sidak this morning claimed it wasn't but that's an unproductive debate, when you want to charge differently for different qualities of service, in a way that isn't simply charging for the increased marginal cost there often can be an incentive deliberately to degrade the low quality service. The industrial organization book, quotes the classic example of the French railroads that spent money to rip roofs off some railroad cars so they could sell really unpleasant service to poor people. So yes, there is a possibility of that. It's part of the conflicting forces, and we don't necessarily know which direction it goes.

>>MICHAEL SALINGER
Anyone else want to pick up on that?

>>MALE SPEAKER
You weren't suggesting that the networks are going to get built by volunteers?

>>JOSEPH FARRELL
No. I'm saying both matter.

>>MICHAEL SALINGER
If we take a wait and see approach, given that the technology is advancing, how would we know whether there has been a phenomenon of degrading the current slow approach in order to be able to charge more for the fast approach?

>>JOSEPH FARRELL
I suppose you could do an investigation and find some smoking documents or something. But I think in general it would be very difficult to know. Yeah.

>>MALE SPEAKER
I think it would be very difficult to know other than doing, you know, some investigating and really, it's a question for the likes of Jon Peha, the people who look at the protocols inside the router.

>>JOSEPH FARRELL
I'm not sure that's where the complaints are going to come. We haven't heard a lot about how the network neutrality concern, in order to do price discrimination, the networks will keep the ordinary quality of service low. It's a possibility, as I just finished saying, but I don't think that's what people are mostly worried about.

>>MALE SPEAKER
I was going to say, Joe is right. There is the whole literature about, you know, versioning and reducing the quality of one good to make sure people buy the high version, is the solution to that, saying, no, you can't have didn't versions? That's like the idea of saying can't charge for higher speed access, would probably be substantially worse than this worry about the lower end. Again, this degrading the lower end becomes less of a problem the more competition you have for access.

>>MALE SPEAKER
Yeah.

>>MALE SPEAKER
That's the key to this question is thinking about that issue.

>>JOSEPH FARRELL
Yeah, I think some of the issues we've been discussing, I'm not convinced that moderate access competition of the kind we have or are likely to get will solve it but on the one hand I suspect it would.

>>MALE SPEAKER
Simon, the major contrast between the shares that you have to contribute for access in these countries, is that simply a description of inadequacy of competition in the United States?

>>SIMON WILKIE
That's a really good question. I couldn't get a good answer when we were asking people. Because we sort of have the same market structure, right? You have different countries with similar numbers of players, and the prices, you know, the rent extraction factor varies from 50 to zero. I think the low numbers in Asia have more to do with the sort of tacit bullying nature of the governments, rather than being at a different equilibrium. That's the best I can come up with. The difference is, they also have more spectrum. The interesting thing, I guess, T- Mobile in Europe, they are moving towards open access, or essentially zero termination model so it might be they have sufficient spectrum for somebody to break the equilibrium there. That's my conjecture. It's just a conjecture.

>>MICHAEL SALINGER
Another question from the audience. If all broadband (inaudible) -- why not have community ownership?

>>MALE SPEAKER
I must say I do not object in principle to communities providing their own facilities. It's a form of competition. I've remarked in the past, a differential taxation made competition unequal. But it's just another form of competitive entry from my standpoint. I do not regard it with disgust.

>>MALE SPEAKER
I think there are two forms of the question. Maybe it's worth clarifying. I think Fred was addressing why not have community ownership of one network, one access provider. In addition, perhaps, to some private ones. I'm a big fan of that. People talk about what a pain it is to compete against the government. At some level, I think, the private sector should say bring it on, and we're always talking about how inefficient the government is. Let's prove it by beating you. The caveat is, if the government provision comes along with a completely bottomless supply of willingness to incur losses, then you can have trouble. The other form of the question, which is actually the way I was tempted to interpret it, is why have private ownership at all, why not just have community provision. I think that at least in this country seems more likely to lead to very serious failure of network investment. And perhaps a failure of helpful imagination in how the run the network so I wouldn't be in favor of that. I'm in favor of diverse competition, and I think government provision is a legitimate part of that.

>>MALE SPEAKER
Where are you on that? (Laughter)

>>MALE SPEAKER
The fact is the studies that have been done, pretty much without exception, of these publicly owned municipally owned telecom networks show that they are a terrible deal for the taxpayers who are forced to support them. They go in typically, you know, with the rationale that, you know, these services are not being provided by the private sector. And that's generally not true. Which, typically there is, and when you are talking about, you know, the fiber, the ones that are laying a lot of fiber, which is very expensive, I think without exception, they just lose lots and lots of money, which the taxpayers or the electricity rate payers, because sometimes these are connected with electrical systems, essentially pay. And, you know, and citizens are essentially forced to be involuntary shareholders in a bad business. The wireless ones, I think, are in a sense are too new but I don't think any of them have been successful except in very small areas.

>>MALE SPEAKER
On the wireless one, one of the problems is the cities are run like a megalopolis. San Francisco, we'll allow one WiFi service to cover the cities. We're going to lease our light poles to one provider. We'll do a deal and that deal will have all sorts of sweet deals for city hall and other things like that. It's a bad idea. It's not serving the citizens, it's serving the elected officials. What they should do is say, we like the idea of people providing networks, using the city's facilities, but we'll allow anybody to use the city facilities and have more competition. If cities do want to put in their own systems, think I, you know, I think it should be allowed. I'm glad I'm not in a city, that I'm paying taxes to a city is doing it, but for private companies to object to a city coming in, would be a bad thing.

>>MALE SPEAKER
If we look at past technologies for these issues, I mean, the relationship between delivery and content, the argument was, well, you need to allow -- recognize the potential dangers of integration, but you need to allow investment in content by the delivery companies because, because otherwise, the content is not going to appear. Or it's not going to be as good. With the internet, it seems like there is ample supply of content. So what would be wrong with saying, you're either in delivery or you're in content but not both.

>>MALE SPEAKER
I don't think that's stupid idea. Going back to what I was saying a few minutes ago, an access provider, a delivery provider, does have additional incentives to innovate in content, because they capture the complimentary revenues. On the other hand, once you allow the access provider to be vertically integrated into something that almost by definition is going to be profitable at the margin, you create incentives for them to either exclude or charge in a way that's going to be very difficult to keep aside from ex-appropriation, rivals in the business. I'm actually -- I would say the vertical separation model is worth exploring. Let me step back. Where are we on this? I think this topic has been debated, as Greg said, since 18 something. I'm -- I'm aware of it, since 2002. I have to say, the quality of debate that I've seen has been abysmal. I hope we can start to debate things at a somewhat higher level and it seems like we've lost five years or something in not doing that. Vertical separation, I think could, be part of the discussion, as could some of these other things. I think we should get going on a high-level attempt to debate this and get away from the, oh, it's terrible. Oh, there is no possible problem level of debate.

>>MALE SPEAKER
I have a feeling that you're not going to be a fan of vertical --

>>MALE SPEAKER
I mean, first of all, I don't think there is, you know, literature on the results of the various vertical separation schemes, I don't think, is necessarily supports doing it, especially for something that's so important as the internet. The other thing is, you know, I'll go back to the fact this is a pretty young industry. Exactly what, you know, to freeze very complicated and fast moving technology, to freeze a particular structure in place now just seemed to me to be when you don't know what the thing will look like in five years, just it seems to me to be extremely risky? I certainly agree with Tom. Partly out of ignorance but I think of the times when we had the rules, the financial interest and central -- syndication rules prohibiting broadcasters from having an interest in programming, the trend has been away from that. Particularly in the situation of innovation. To have vertical separation at this time, I want to respond to it. I would be opposed to freezing the structure.

>>MALE SPEAKER
There are lots of efficiencies for vertical integration that could arise. When you say ample supply of content on the internet, it's true, there is a lot of stuff out there but it may not be the right stuff that people want to use, for example, that may cause people to increase their demand for broadband even though it may be a zero profit on the content side. So there are all sorts of relationships that can improve efficiency by having vertical relationships. So I think ruling it out is probably a bad idea. There are the fears of these ideas of ex-appropriation or other things that may come about. And you should worry about those. But really not having vertical integration is probably a bad idea at this point in time.

>>MICHAEL SALINGER
What would raise the level of debate?

>>JOSEPH FARRELL
I think for example, just talking about vertical separation, what have been the success stories, what are the failure stories, what do they have in common? I'm interested but in that subject but I don't think I could give you a good answer to that question. I don't think the debate on net neutrality has contained any good exhibitions on the answer to that question. Can I come back in just a moment on vertical integration? The economic models that say vertical integration helps are by and large models that operate at the level of incentives. Incentives are important for innovation, but as I was starting to say a few minutes ago, one of the lessons from the flourishing of the internet is incentives are not the only thing that's important for innovation. As Fred was saying there has been a trend away from vertical separation, largely driven by increased attention to these incentive issues. I think we want to be very careful not to throw the baby out with the bath water and lose track of the fact that although incentives are very important they are not the only thing that's very important here.

>>MALE SPEAKER
(inaudible)

>>MALE SPEAKER
I got confused about who was the baby and who was the bath water? (laughter)

>>MALE SPEAKER
Let me ask a question, would it be a bad thing if a network operator wanted to start a search engine?

>>MALE SPEAKER
Wanted to start surcharging for what?

>>MALE SPEAKER
A search engine.

>>MALE SPEAKER
Oh, a search engine. Well, let me deal with the easy case first. If the network operator wanted to start a search engine, and not accompany that by either blocking access to Google or saying to Google, we now have a private opportunity cost, also known as a lost profits component of private costs of dealing with you, we're going to charge you some allegedly ECPR like amount of money. If neither of those things happened and they just started a search engine, that would be fine. If they did one of those things that would not necessarily be bad. What that would do is that would enable them to set a pricing structure in which, at some level, whether by money, ads or something, they were charging for the use of the search engine that would enable them to do a more complicate price structure than they would have otherwise. That could help with the second best pricing problem. It would move us further away from first best pricing. And it would also raise, I think, a variety of concerns that wouldn't necessarily be a big problem but that might easily be a problem having to do with, well, what are they really doing to incentives of independent content providers? That's where I think it would get difficult.

>>MICHAEL SALINGER
We have five minutes left and I want to give people a chance to put in a final word, go from left to right. My left to my right.

>>MALE SPEAKER
Am I on the far left of the panel?

>>MICHAEL SALINGER
Yes, you are. Physically you are.

>>MALE SPEAKER
I think there are, you know, my view of this there are concerns about what a firm -- an ISP with market power can do. The most important thing, I think, is the FCC should get more spectrum in the marketplace to try to ensure there are multiple providers of high-speed internet access to consumers so that consumers have the choice, and that will help to discipline a lot of the problems that we've talked about today.

>>MALE SPEAKER
Partially just to echo Greg's comments but to also add I think, you know, to a big extent there is real issue here. Not so much the discrimination issue that's been talked about in terms of speed since tiering is a sensible market approach but terminating monopoly problem. The problem of finding a connection is real. We have something already in place that deals with it. So just proceed with prudence and caution.

>>MALE SPEAKER
Yeah, I would stress that the aspect that Fred was talking about and the relative youth of the industry, it is a young, rapidly changing dynamic industry. Even though things are not perfect it's hard to really see how [inaudible] would make it better.

>>MALE SPEAKER
By chance, I have a close analogy between my reference to immediate medical diagnosis, access to records diagnosis and treatment, requiring very high speed access, and to a medical problem. Above all else, do no harm. I think that applies to premature efforts to jump in and regulate this industry.

>>MICHAEL SALINGER
Joe, you get the final word.

>>JOSEPH FARRELL
Thank you. I think I led off by saying you may not be able to tell where I come out. And actually, probably five years ago now, Tom Lenard organized a conference on this, where he said he wanted to put together a balanced panel and I offered to be a balanced panel all by myself. I'm very well aware of, I think most, certainly a reasonably representative sample of the arguments on both sides. I am certainly well alert to the dangers of regulations. I'm well aware of the importance of providing good incentives for network and well as content investment and innovation. It's very easy to be the famous two-handed economist. I'm pretty good at that actually. I thought it would be a little irresponsible not to let you know what I would say if I were waking up at 3:00 in the morning and asked to make an immediate decision. So that's what I would like to leave you with. Being aware, as I am, of all -- or of many good, sound serious arguments on both side, as a consumer, I would regard it as very worrisome if I woke up one morning and there was AT&T or Comcast plunging into the kind integration and negotiation that we've been talking about. I have to disclose that while that worry is informed by my professional expertise, not implied by my professional expertise, it's perfectly possible to be an expert economist and not worry about that, but to be honest, I worry about it, and so I would like to feel that there is some kind of protection against that kind of thing happening, and against the Internet becoming Vulcanized, whether or not that would be profitable, whether or not it's a very likely concern, I would like to feel protected against it.

>>MICHAEL SALLINGER
Thank you, to the panelists. (applause)