Broadband Policy: Do We Really Need Legislation?

Transcript

MR. ROBERT W. HAHN: AEI-Brookings Joint Center. We'd like to welcome you.

For those of you who don't know about us you can find out a little bit at our web page at AEI.Brookings.org. But our basic mission is to hold lawmakers and regulators more accountable for some of the things that they do as in decision-making.

We also try to educate the public and ourselves about important issues of the day like broadband, the one that will be discussed here.

Before we get to the debate I want to give you an advertisement. In your packet you have a blue card. If you want to be kept up to date on the latest and greatest happenings in Washington, D.C. please fill out the blue card and give it to one of the nice young people at the desk outside.

Speaking of that I'd like to give special thanks to the Brookings Communications Department for helping to put on this event and in particular Robert Dabrowski and Marcia MacNeil.

Additionally one more ad and then we'll get started, or maybe two more. We keep all our publications on our web site so they're almost free for the asking. And Tom Hazlett would be particularly appreciative if you download it, click on his so he can be in the top position every month when we issue our top ten report.

The Joint Center is in the process of publishing a book on broadband which should be out just after the political debate is over. (Laughter) No, actually the book is being edited by Bob Crandall and it's got very diverse perspectives in it and I think it represents some very good scholarship on this subject.

In addition, one final ad, Dr. Ferguson who is one of our speakers today also has a book coming out on broadband which will be published by Brookings, actually it's ready to be published.

I'm told there's going to be a pre-book signing in the lobby after this event for those of you who are interested.

If you just wandered into the room by mistake let me start with a lay person's definition of broadband and Tom and Charles can expand or deny or do whatever they want to with. Broadband technology basically allows for high-speed access to the internet through services like DSL, cable, and satellite if you can find it. It would make certain applications like streaming video much better and I'm told more practical, and it could even allow me to watch video on demand one day when I wake up hopefully in the not too distant future.

As I said this morning we're very happy to bring you a debate on broadband and the debate's title is "Broadband Policy: Do We Really Need Legislation?" Unfortunately I think that title is a bit of a misnomer and I was hoping Bob, that you made the title so I wasn't criticizing myself.

MR. ROBERT E. LITAN: I did it.

MR. HAHN: Then I apologize for criticizing you. But as most of you know, we already have legislation—the 1996 Telecom Act—which was aimed at opening up local telephone markets to more competition. Whether it's successful or not is another matter and Dr. Hazlett has a very nice paper on our web site which addresses that issue.

I think the real question that we need to ask is whether more legislation is needed and if so, what form should it take.

As you probably know there are very powerful interests representing both sides of the debate. On one side are the cable companies and the competitive local exchange carriers; on the other side you have folks like the Bell Operating Companies.

The debate is more than academic. There are a number of pieces of legislation on the Hill and the FCC is also going to likely get into the act. Perhaps the best publicized bill is the Tauzin-Dingell Bill which would exempt from regulation certain advanced services popularly know as broadband and in so doing it would lift some important restrictions on the Bells.

The bill recently passed the House but we don't know what's going to happen in the Senate, at least I don't know.

In the Senate, Senators Hollings and Stephens have proposed their own bill which would increase oversight of the Bell Companies.

There are several interesting economic questions that the broadband issue raises and I hope our discussants will touch on some if not all of these.

First of all, what's all the fuss about? Is broadband a big deal or not a big deal to consumers as you see it?

Secondly, a related question, would broadband penetration be substantially different if the Baby Bells did not face regulation in this market?

Third, a political economy question, where are we likely to end up in terms of legislation and regulation?

Based on my experience with titans like the cable companies and the Baby Bells when you get out in the legislative arena significant legislation is unlikely unless there is a perceived crisis. But I could be wrong about that. I'm not sure people perceive a crisis here either.

And if a political deal is struck and we do get more broadband legislation in a different regulatory structure, who are the likely winners and losers?

Now that I've posed a few of the questions and taken up more time than I should I'll leave it to the experts who are far more knowledgeable than I to hopefully address some of these issues.

With that I'd like to turn it over to Bob Litan to lay down the ground rules for our session. Thank you.

MR. LITAN: Thank you, I'll be short and to the point.

We have two outstanding debaters today both of whom are friends of mine, and I hope they remain so after the event.

On one side, to my far right—far right only geographically—we have Dr. Charles Ferguson. Charles has a very distinguished and varied background. He used to teach political science at MIT. He is the author of "Microsoft Front Page" if you've ever used that program. And went into business and has done a lot of different things in his life. Then he wrote a book which I highly recommend to you called "High Stakes, No Prisoners" which, and I'm not embarrassing him, and I've recommended it to many people. It's the single best book I have ever read on the venture capital industry and the high tech industry in the popular media. I highly recommend it.

He's now working on a book on broadband, hence the reason he is here. I assume he's going to take a position opposite to our other expert, Dr. Tom Hazlett. Tom used to teach at the University of California Davis. He has an affiliation with the Joint Center, the AEI-Brookings Joint Center, and he is now with the Manhattan Institute.

Tom is a very witty, very engaging economist who will tell you what he thinks in very short order and you will discover that. Those are our two protagonists, and I will just close this introduction by saying this is the public policy equivalent of the "Thrilla? in Manila", so hold your seats and we're going to start with roughly 15 minutes on each side. We'll give brief rebuttals. Then all of you can join the fray and let's see what happens.

Over to you, Tom. You're going to go first.

MR. THOMAS HAZLETT: That's very appropriate, Bob. Thanks very much by the way for that wonderful introduction. The author of the best book ever written and a funny guy talking about broadband. (Laughter)

MR. CHARLES FERGUSON: Sounds good to me.

MR. HAZLETT: guess we can pretty much wrap it up from there, but— (Laughter) Anyway, what, I have 15 minutes for jokes and then we can get to—(Laughter)

Some people seriously said how do they let you debate this topic, and I said look, Tyson's getting a license to box in Washington. I mean it's the only place they'll let me do it. Come on! (Laughter) Anyway, you've got nice ears, but—(Laughter) We'll keep them that way, I'm sure.

I'm glad that Bob Hahn has already attacked this terrible title. I wanted to know who had ownership to this. Do we really need broadband legislation? Well, this is broadband legislation, hoax of fallacy on the first pass. Not only is it a loaded title, I don't know what a yes means and I don't know what a no means, but I think it means we're supposed to talk about broadband policy so we'll take it from there.

I was having a wonderful parenting moment the beginning of December when our three year old, going to be four at the end of this month, actually was, my wife and I were both there watching with the two kids, two little girls. The three almost four year old actually, voluntarily shared something with her two year old sister. This is one of those breakthrough episodes, a real opportunity to give the parenting moment. So I jumped in and gave her the lecture about reciprocal dealing and long term financial self interest and how this was a very rational move for her to discover that she had made her little two year old sister happy about this, and really trying to make this point.

So in sort of in summary I wanted to see if she had gotten this. So I asked her, I said so you understand about sharing? She looked at me and said, "Daddy it's an idea that comes into my head every day." So I thought that's sort of like the FCC. The sharing idea is an idea that comes into their head every day. They just really don't know what to do with it after that point because it's a fairly complex thing, this sharing idea, and where you take it from there, who knows?

I'm going to spend my few minutes here really talking about the network sharing idea. It may not be exactly what the title has in mind but I think this is really where the arguments come. How much network sharing should there be? I think everybody agrees there should be some mandated interconnection between physical networks.

So taking it from there the question is how much more mandated interconnection or network unbundling, how much and how far should it go?

We really I think need to approach this from a very common sense perspective. Mandated network sharing really does spring from this idea, and it goes back certainly before the '96 Telecommunications Act, but just taking it from that point forward, there certainly was an idea that in the long run facilities based competition, everybody agrees that's the thing we want. It's going to be really a lot better if the regulators don't have to get involved figuring out the terms of the mandate and the network sharing. No telling how much real competition network sharing actually creates.

So in the long run we all want facilities based competition. In the short run, however, it may be tough to get a lot of new facilities built and really competing. And in the short run, i.e. the life of a political career, we really do place a lot of weight on getting entrants into the market and at least looking like they're offering some competitive choice. There may be some good economics to this, there may not. We can talk about that and certainly will.

But the basic idea certainly of these network sharing mandates is to impose some sort of regime that will improve the rental option so that the new entrant does not have to go to the expense, trouble, and scale whatever barriers are involved in actually creating new facilities.

And the real crux of this, to simplify greatly, but to really catch the crux of it is that the government really has to concern itself with the terms and conditions of access more viable to the incumbent network. That imposes wholesale price controls, how much the legacy network can charge new firms to use all or part of its system to serve customers. That whole regime of wholesale price controls has obviously been rather exciting over the last several years under the terms and conditions generally laid out in the '96 Act, and then implemented not by the FCC, although the guidelines are established by the FCC and the models and the general logic, but actually implemented by the 50 states.

In putting these rules together that change the rent or own decision by the new competitors in the market, the tenants, the CLEC or the new entrant in the market certainly is affected in its incentive. It has to make its rent or own decision based upon these regulated prices.

Now of course the logic of the network sharing system is to force down the incumbent prices from what would be a monopoly level, would be inefficiently high and would discourage people from utilizing the network. So we're really trying to get more network utilization of existing facilities, vetting the monopoly problem, but of course in doing that we, in fact I'm not surprised, we're actually trying to get the entrants to not build systems, not build competitive facilities-based systems, but to actually share what's actually in the market right now.

Now one of the things we've learned the very hard way in this boom and bust cycle in the industry is that the terms of the unbundling, the pries set in this wholesale rate regulations scheme cannot be used to effectively subsidize the entrants. Why is that?

Well because if you make the prices nice and low and attractive for the entrant, entrants will come in, and lots of entrants will come in, and you do not achieve a stable equilibrium. In fact it's over similar to the problems you have subsidizing much of the agricultural sector. When you throw the subsidies out in fact you get too much production and too much entry, land prices go up, land prices are a fixed asset, not perfectly inelastic but some degree of inelasticity. Land prices go up and farm acquisition costs rise.

We've seen in the competitive local exchange space customer acquisition costs actually rose to about twice network utilization costs in the episode just concluded, the boom phase when we got all this entry at regulated wholesale rates.

So the effort to actually subsidize entry is doomed to failure.

Of course the landlord, the owner of the legacy system, be it a monopoly or not, has a lot of incentive changes with the mandated network sharing. First of all now the owner of the system has to face the fact that any improvements, any upgrades that it performs has to be done in the light of how it affects the competitive stance of the company vis-a-vis its rivals, and its fiduciary shareholders, the fiduciary responsibility to shareholders, the owners of the legacy system, is not to help the rivals more than the actual owners. They'd be subject, of course, to shareholder lawsuits if they did.

Part of this scheme, of course, revolves, this wholesale rate regulation, is that the owner of the legacy or the regulated system that has to open it up at wholesale regulated rates, actually supplies options. That is to say the regulated rates are not guaranteed. I should say the rate is guaranteed but the revenues are not guaranteed. That is to say if the investments are made by the regulated network owner and the market ends up looking good for entry, there will be some entry, there will be some utilization by the competitors or rivals who use the system but there's no guarantee of that. In fact if the market goes south you can see a situation like today, that there is an exit of much of the wholesale utilization of these systems. And in fact the revenue does not materialize. So it's a situation, you've heard the expression "heads we win, tails you lose", but the cost of these options really is a disincentive on investment in infrastructure. And at a time when the technology is dynamic and there is a crying need for more investment in broadband and we may well agree on that, the cost of the options that are written by the network owners tends to be a very large disincentive.

People talk about the fact that regulators are likely to set the wrong prices. Now of course that's true, it's just hard to get them right. They're either too low, in which case there's over-entries and over-utilization of the incumbent's network and too much discouragement of competitive facilities, or they're too high and there's too much duplicative investment. That's the normal view. But the idea here that when you take risk into account is that these systems, all these new broadband systems are built as lottery tickets where you don't really know where the revenues are coming from. And to mandate unbundling at cost-based wholesale rates at any price tends to be a tax because of the cost of the option being written by the investor, it tends to be a tax on the investment. That stems from the fact that revenues will only materialize if in fact those rates that are set by the regulators turn out to be good deals in the future for the entrants.

And of course this takes you to a situation where the unbundling operations themselves create a free riding situation where the investor has to anticipate appropriation of their risky investment in the event the investment turns out to be quite profitable. Meaning the customers like the services that are provided. And this is absolutely irrespective to whether or not it's an entrant or a legacy monopolist that's doing the investing. And this is just the quotation from Cox Cable which would be defined in the telephone space as an entrant or a CLEC, Competitive Local Exchange Carrier, and it talks about the fact that just from a 10K Securities and Exchange Commission filing in the year 2000 Cox notes to investors that they have requirements, unbundled systems, and that places them in a situation of basically subsidizing their competitors.

So entrants see this and in fact investors see this. In a piece for George Bittlingmeyer I wrote last year—I know this is factually the other way around, Bittlingmeyer & Hazlett, but I'm sure that's a subconscious juxtaposition there. (Laughter) The fact is that if you trace cable open access regulatory movements, when cable open access attempts to move forward you do not get a positive response from the shareholders and investors in content and internet infrastructure firms. And when in fact cable open access is dealt a setback in the regulatory environment, and there have been a lot more setbacks, the three events we can identify where open access and cable have lost, then you actually get an increase in the value of internet content and internet infrastructure firms. So we're not looking at site at home here or the cable companies themselves that are the direct objects of regulation, but by the providers of complementary inputs and that's a very good metric on what's going on.

One way to look at this is just to see the situation as we have it out in the marketplace today. There is a race on. A broadband deployment race. And we have two leading technologies. Hopefully we'll have more leading technologies soon, and deregulation, in fact, whether it be through the legislative or the regulatory process would help on that, but the race that is on today shows us something important about the impediments of regulation.

A closed platform is winning the race in the United States, not only in the United States but internationally. We can talk more about that later perhaps.

In the United States DSL, which came on the scene first but didn't deploy for quite awhile, finally deployed when cable modem service and high speed access on arrival platform started taking off, was supposed to do very well. If you go back to 1998, 1999, you see important voices within the tech community including John Chambers, Cisco Systems, saying that DSL's a better system, it's going to win. But in fact the modular delivery system that's mandated for DSL because of the unbundling requirement of the underlying local exchange carriers is quite distinct from the regulatory regime in cable where you have a vertically integrated monopolist in the cable space providing what may or may not be a competitive service in the broadband space depending on how you want to identify this broadband competition between DSL and cable. But on the monopoly cable platform with the vertical integrated system you have clear leadership in the race in the United States. Now it's not that way in all countries, interestingly enough, and where there's fewer unbundling requirements it turns out there's actually more deployment of

DSL relative to cable in an international analysis.

One of the very interesting outcomes of this is that not only is cable typically leading the race, but one of the reasons given by almost all observers on this is because cable is a closed system and they have greater integrity through vertical integration, control of their product not only in a service sense but also in terms of creating standards for the equipment manufacturers. This is one of the great ironies, is that there are open standards that the cable industry has established for cable modems. That has been a big, big encouragement to competition in the equipment manufacturing space to help the cable industry with its vertically integrated monopolies actually do one better than the DSL space with its modularity imposed by unbundling mandates.

So this is what the trade press says, the DSL providers are getting their butts kicked in the residential market. I suppose that's a technical term. It's actually a web site that's very favorable to DSL in general, in fact the DSL industry. But they talk about the advantages of this vertical integration and even the open platform.

There's obviously a lot we ought to talk about. I'll just conclude by saying that I think in going through this and many many other examples from retail rate regulation in cable to video dial tone and open video systems which is a common carrier approach, to video competition and the failure of that program, I think that you really see here that with the beginning of facilities-based competition and in fact some strong competitors coming on including satellite and fixed wireless, that the clear policy implication is that we should go forward with a deregulatory program that allows these new players to compete very very effectively with the established cable modem and DSL providers, but that we take the mandates off of, to some degree, the DSL providers. That is to say the Tauzin-Dingell approach to releasing the mandates, the unbundling mandates for new investment is absolutely appropriate, as well as the FCC's recent approach and ongoing approach to in essence blocking open access requirements in cable. I think that that is working in the marketplace today and is actually quite a solid success in terms of its non-regulatory approach.

I'll end there.

MR. FERGUSON: My slides and my jokes aren't going to be as good, but here goes.

Let me start by kind of backing up a bit and trying to say something about what the stakes here are. The two of us seem to agree on that. I'm not quite sure how much we agree, but let me just make it clear what my own view about what the size of the stakes is. I think the stakes are very large.

This is one of those cases where something that is somewhat undramatic and not terribly sexy, doesn't involve a lot of good video images except in perhaps the delivery product in the end to your television screen or your computer, it turns out to be macroeconomically really extraordinarily significant. The significance derives in part from the fact that what's important here is not only delivery of broadband services to residences but also delivery of broadband services to businesses which is something that's often forgotten in the discussion of this question.

Let me just kind of give you some numbers very quickly.

Broadband itself is a tiny industry. By the widest possible definition, broadband revenues in the United States are maybe $25-$30 billion this year, something like that. That includes things like T-1 service to businesses, it includes business services. The residential broadband market is maybe $5 or $10 billion at most. So you might think it's a small industry.

Actually what's important about broadband is that it offers the opportunity and also the threat, which is what's going on here, to transform very large industries in a very deep way. Local telecommunications, including voice services in the United States alone is about $150 billion this year. And including long distance telecommunications and the cable industry, telecommunications in the United States is $250 to $300 billion. That's services. That doesn't include equipment. Then you add in the equipment and the equipment includes personal computers to talk over the internet these days, then you're up to kind of $500, $600, $700 billion per year in the United States. Then the really big question, although already $600 billion a year is kind of a big question, but the really big question is what role does the internet, what role do information services play in the economic growth of the United States and the economic growth of the world.

It's very hard to tell. I have a constitutional and also fairly learned distrust for all econometric estimates of these questions. I've looked at most of the productivity studies and I sort of don't believe them, but nonetheless if you pretend that you believe them then it appears that the productivity revival in the United States owes something like one percent per year to the progress of information technology and/or the internet since the early to mid 1990s. There's a lot of dispute about that. Some people think it's more, some people think it's less, there are huge measurement questions about what's productivity and what's information technology and so on. But I would take as more reliable the view of anybody who has swum in this ocean, you just know this is very very important.

So why did I say all that? I said all that because if you look at the whole world of information technology—semiconductors, computers, telecommunications equipment, private corporate networks, integration services, consumer electronics, it's a very large industry, and you observe that almost all of this very large kind of $1 trillion per year industry is changing technologically at 40, 50, 60 percent per year. Significant portions of it, by the way, the fastest moving portions of it being in communications, are changing technologically at up to 100 percent per year. That measurement is the price performance of delivered products and services. So how many dollars per byte per second for communications bandwidth? How many dollars per mips, for processing power in microprocessors, things like that.

There is one staggering exception to this rule and that staggering exception is local telecommunications.

So if you look at telephone service, voice service; if you look at T-1 service which is 1.5 megabytes per second data services; even if you look at DSL. In a second I'll do a very interesting, for me anyway, calculation about whether DSL really represents fast, impressive technological change. The answer is it doesn't.

What you find is that the local telecommunications industry in the United States for a very long time has been the only portion of the digital information technology sector that has been changing, technologically speaking, again, delivered price performance, maybe 10 percent per year, in some cases zero, in some cases even negative technological change, which is really quite an extraordinary achievement when your underlying technology is improving 50 percent per year. You have to try very hard to get zero technological improvement for a decade. It takes a great deal of effort to do that. But our local telephone companies have succeeded through their enormous efforts.

I sat through IBM in the 1980s and '90s and it's very reminiscent of IBM and indeed the regulatory structure matters somewhat, but there are a lot of ways in which it turns out not to matter very much.

One way in which it's reminiscent of IBM is that during the period that IBM was going down the drain in the 1980s and early '90s, culminating in its losing $30 billion in 1993 and then firing its CEO and then firing 200,000 people over those next 18 months, during the time that was occurring IBM every year had R&D expenditures and also usually had profits which were larger than the total revenues of its largest competitors.

So when you have a very comfortable, cozy, sloppy monopoly, money ceases to be an issue in the usual ways. In fact it kind of begins working against you in terms of your competence and abilities because you have so much of it that you can paper over enormous problems until finally those problems come to roost.

If you hold the view that what should happen to local telecommunications is that it should become a Silicon Valley- like, very fast moving, standards driven, extremely competitive, dynamic industry displaying very rapid technological progress, then you are pretty much driven to the view that you have to do something very, very dramatic to the current structure.

The Telecommunications Act of 1996 was in some ways kind of a noble attempt. What frustrated it I think was a combination of the lack of technological competence of and technological expertise of the regulators in the government, although some very gifted people went into the FCC—my friend Joe Farrell was the chief economist at the FCC for awhile, a very smart guy. But in the Justice Department, the Antitrust Division, for example there's 350 attorneys, 50 economists. How many technologists? The answer is zero. Literally. There is no chief engineer, there's no chief technologist at the Antitrust Division in the Justice Department. There is a chief economist.

So that's one problem, is the government doesn't know how to deal with these kinds of industries and that shows up in many different ways. It shows up in the courts, in the amount of time that the antitrust proceedings take. It shows up in the regulators and regulations. The Commissioners of the FCC are not technologically sophisticated people. They generally come from the broadcast industry or places like that. Or else they're just people who were lawyers and good fundraisers for somebody who became President, or the son of the Secretary of State. (Laughter) Just to mention a random possibility.

So that's one class of problems.

The other class of problems is that when you look at the telephone companies, what is their true core competence? Their true core competence is lobbying. They're really good at it. They spend much more on it every year than they do on R&D. That is literally true.

I've been trying to make estimates of how much the ILECs spend per year on lobbying in all forums, and it's very hard to tell, but a reasonable guess is somewhere around $300 to $400 million per year. Very serious money. That's composed of paying lobbyists, it's composed of their Washington, DC government affairs offices, it's composed of their political contributions, it's composed of the many eminent academic experts they hire to write testimony and regulatory pleadings for them and academic papers on behalf of their cause. It's composed of all the people they have in state government, in state capitols doing the same kind of regulatory lobbying. It's composed of lawyers they pay to file lawsuits against the FCC and the government and so forth.

When you add all that up it turns out to be really an enormous amount of money.

To conclude, what do I think ought to be done about this situation?

First of all there are some background things that are to some extent independent of the particular industry which would be very helpful, campaign finance reform being one of them. A substantial increase in the technical capabilities of the FCC, the Justice Department and other parts of the government that are relevant here would certainly be another.

But the $64,000 question is what do you do about the telecommunications industry including, by the way, the cable companies who have the virtue of being the challengers at the moment and therefore escaping the kind of scrutiny that otherwise they would deserve. They're not going to be our saviors, both for incentive structure reasons and also for technological reasons which I can go into later if you're interested. The telephone companies are the big problem.

If it were politically, legally, logistically and otherwise feasible, I think what you'd want to do would be to cause a fairly major structural divestiture which would not necessarily solve the entire problem. There is a residual problem which one cannot deny that certain parts of these systems may, I think it's very open whether this is true or not, but there's at least a reasonable possibility that certain parts of these systems are approaching natural monopoly status in the amount of money that it takes to dig trenches and put fiberoptic cable into them. The problem being the digging of the trench, by the way. The cost of delivering fiber to somebody, whether it's a home or a business, it doesn't matter, is about 80 percent digging the hole and filling the hole in. It's not the fiber that goes into the hole.

So in addition to some kind of very tough structural divestiture, which would really be a very major landmark antitrust event comparable in size, scope and I think economic importance, perhaps larger in economic importance than the divestiture of Standard Oil early in the 20th century, in addition to that I think that there is a good case for tax incentives and significant financial incentives for the construction of bandwidth, for the physical delivery of bandwidth. Those incentives do not need to be and should not be given primarily or beneficially or disproportionately to the incumbents. There are, and this is kind of the good news in this, there are other parties who are potentially willing and able to take advantage of such incentives. Those include owners of very large buildings, large real estate owners, they include municipal governments, they include non-profit organizations such as universities, they include the government.

So I think there is the prospect, if the regulatory structural situation were changed significantly, I think there is a prospect of liberating a lot of economic activity and appropriate tax incentives which would be difficult to design in miniature, but appropriate tax incentives would potentially be a significant and beneficial piece of the overall solution, and I think the one respect in which the things that the telephone companies say have some justice to them, that is that doing this is going to be over time a fairly expensive proposition.

With that I will conclude.

MR. LITAN: Do you want to have any brief comments on this or do you want to go directly to Q&A?

MR. HAZLETT: I'm certainly glad to hear that Charles favors structural separation because I was afraid we were going to agree.

This is a big deal. Broadband is very very important. And a regulated structure is not consistent with getting modern communications out into the marketplace.

The attack on the LECs is silly if taken as an attack on the executives and shareholders of the LECs, and I think it's quite appropriate if taken as a criticism of the highly regulated system that is incompatible with the risk-taking and the dynamic volatilities that are apparent in the markets today in adopting new technology. I think that's certainly where we might agree.

There is another market in which zero percent productivity or less than ten percent annual productivity is certainly observed, that is in the delivery of local cable television services.

I'm not going to downplay the important investment, large-scale investment by cable investors and in fact improvements in cable television over the last several years, particularly with deregulation of rates. Rates have actually gone up less since the deregulation of March 1999 on a real basis, inflation adjusted, than they went up over the period of about seven years of rate regulation under the '92 Cable Act.

But the fact is that in the local delivery of services you're not seeing the tremendous 1500 percent annual productivity gains. I don't know if chastising one particular set of producers is the way to go about looking at this. I think if you're going to look at the local producers of telecommunications access, and it's quite interesting. Compare the local cable systems without rate regulation essentially, without unbundling obligations to the LECs.

The cable operators, to the extent that the Bells are successful as lobbyists—The fact that you spend a lot of money does not mean it's a source of competency. If you look at the Q ratios in the cable industry you'll find that right now cable systems cost about $4,000 per subscriber to buy in the marketplace. They're actually selling it for $4500 per subscriber today. That's a contest. They cost about $1,000.

All the capital that goes into them, $1,000 brand new.

On the ILEC side you get a Q ratio—That's a Q ratio of over four, the ratio to market value to the capital costs, and that's a measure of the projection of the super competitive profits that are going to be capitalized by the investor that owns them.

On the ILEC side, you've got a Q ratio of a little over one. You've got a system selling for $3,000 or a little more per line, and they cost about $2500 in terms of all the capital that goes into them according to the projections that I have from Legg Mason.

So the real market power is on the cable side.

Having said that, cable is actually doing a lot in the broadband space to bring services to customers, cleaning the clocks of the DSL competitors, so to speak, and providing digital cable, pay per view, virtual video on demand and other services, cable telephonies, that are rather interesting in terms of advancement.

The last thing I would say on that is in terms of actually bringing competition and choice to customers, the unbundling approach that says that we've got to use the existing monopoly network on a shared basis through hotel price controls and all these kinds of rules, that approach has actually yielded less than about five percent of the residential phone lines in America over the last six years to CLECs.

Today in America, it's probably interesting, you don't hear anything about this, according to the FCC there are approximately 14 million homes, about 14 percent, a little over 100 million U.S. households, there are about 14 million homes that have the opportunity to buy cable telephony or a dial tone service from the local phone company. Now you may not think that the phone companies are doing much of a job. The cable tax rate is only 14 percent. So 86 percent of the public rejects your view that the phone companies are doing a bad job. Now some of that may be the success of regulation. It's somewhat complicated. The rates may be regulated in a way that local telephone service is a pretty good deal.

But the fact is that when an unregulated alternative comes along only about 14 percent of households actually take that option and that's an interesting fact. But the more interesting fact to me is that about 14 or 15 million households today have the choice of one or two local access providers.

So I think that facilities-based competition is already coming into the market, even in the toughest part of the market, a part of the market that's subsidized heavily, local access, and when you look at the broadband markets developing as they are I think that clearly going to deregulation is going to provide the incentive for investment. Structural separation is going to decrease investment for, we'll talk about the many reasons perhaps later.

MR. LITAN: Charles, do you want to have a quick reply?

MR. FERGUSON: Yeah. Let me just address the question of whether the cable industry is our savior or not and whether in the absence of some kind of fairly forceful policy intervention in the natural course of events the cable industry would take care of our problem.

There are about 14 trillion reasons to be extremely skeptical about that. I would love it if it were true. I don't particularly enjoy having this problem. Believe me, I have other things I can do in life besides write books about the broadband question and the people who run Cisco and Intel, believe me, they would rather not have this problem as well and they're getting quite upset.

One Intel executive recently said to me, you know, we make Ferraris. The problem is they have to drive on dirt roads. And he's right.

Actually he said that to me a couple of years ago, in fact.

So let me talk about this question of the cable industry. First of all, I mentioned earlier that it was quite important that this was a business problem as well as a residential problem. Cable systems don't pass most businesses in the United States. They pass about 90 percent of residences. Cable is only subscribed to by about 70 percent of residences. And cable passes a much lower fraction of businesses. That's problem number one. A small minority of businesses. An enormous investment would be required in order to change that and it would be a very unnatural investment for the cable companies because businesses in general are not large consumers of one-way consumer entertainment. They're not consumers of cable television which is their primary, in the case of the industry, their primary service. So that's a problem.

The second problem is the way cable systems are currently architected and constructed, they are very good for delivery of a lot of stuff to a residence or a business if a business is passed. They are very bad for sending stuff from a residence or a business. To get so-called symmetrical bandwidth you need a lot of investment and very different kinds of architecture.

Very interestingly it turns out the telephone industry is also delivering asymmetrical services, even though the telephone system is, technologically speaking, much more naturally set up to deliver symmetrical bandwidth. Why is the telephone industry delivering asymmetrical bandwidth? Very interesting.

It's a very simple reason. The very simple reason is if you deliver very high bandwidth symmetrically then over a single copper wire you can run about 20 trillion voice conversations which would not be good for the telephone company because the cost of telephone service would go to about six cents. There is a huge cannibalization issue that the telephone industry is avoiding.

The cable industry has a different cannibalization problem. They love delivering what is currently rather amusingly called broadband which is up to maybe a megabyte per second downstream, something like that. You need about 20 megabytes per second to deliver HDTV quality to a residence. The broadcast digital HDTV standard is 19.2 megabytes per second compressed.

Now if you deliver, and it's perfectly technologically feasible to deliver that kind of bandwidth. There's an emerging standard called DDSL that delivers up to 50 megabytes per second. You can get up to, for short distances, up to 250 megabytes per second over a copper wire if you deliver fiber to the pedestal which is—you don't need me to go into all this. But the technology is not the problem.

However, would a cable company which has a great deal of proprietary content which it owns and also makes whatever monopoly profits it makes by virtue of the fact that it controls the only broadband that's capable of delivering a lot of entertainment, would a cable company like the idea that for internet services kind of rates anybody in the entire world could deliver HDTV quality from anywhere to anywhere on demand over the internet? No. This is not an idea that the cable company likes.

So there is a huge incentive structure problem in the cable industry as well.

Finally, with regard to the question of competing with telephone companies, delivering telephone service is a package deal. It doesn't have to be that way technologically, but that's the way it is. The package deal includes not only dial tone but it also includes caller ID, voice mail, call forwarding, three-way calling, customer service, dot, dot, dot, dot, dot.

The idea that cable companies and the cable industry in anything like their current configuration can deliver that package of services in a way that competes effectively with a telephone industry that is stonewalling them at every turn and which could conceivably begin to compete with them in a serious way down the road in broadband services if the two sides did not keep to their respective monopolies, I think one should be extremely skeptical about the idea that you're going to have the cable industry, a $40 billion industry, a third of which comes from content, competing effectively in the telephone market, the voice services market with a $150 billion monopoly incumbent.

MR. LITAN: Here's what I want to do. I want to open it up, and rather than take one question at a time I think we're going to take maybe two or three or so, we'll accumulate them, then do a response and go back.

We have mikes. State who you are and make your point or your question and then we'll go on.

Q: Peter Orszag from Brookings.

Two quick questions. One for both of you. Broadband penetration rates in Canada are about twice what they are in the United States. I was wondering, maybe that's not high enough but it's still dramatically higher than in the U.S. I was wondering if you could speculate about the possible explanation.

Secondly for Mr. Hazlett, one thing you said made me think that you're asserting that no regulated price can capture the option value and I wanted to just check to see whether that was your view. It seems to me like the regulated price could, at least in theory, albeit perhaps more difficult in practice, reflect the real options involved.

Q: —Hopka with the National Retail Federation.

The competitive carriers have been frustrated in their ability to deliver and they say that the ILECs have not delivered on orders, that orders are placed and are then not fulfilled. The ILECs say that they lost them. Sorry. Truckloads of them.

I'm wondering whether our panelists think that the ILECs are evil or incompetent. (Laughter)

MR. LITAN: Can I ask the third question, then we'll go to the panel, just to follow on that one.

To believe or to argue that we ought to deregulate the phone companies on the broadband side you'd have to believe that we're going to get more aggressive competition and more aggressive behavior by the phone companies once they don't have the sharing obligation and once they're a total monopoly. I can just tell you from my own personal experience, and I'd bet you, I'm not going to ask you how many of you have DSL from our local telephone company, but my experience was sheer hell with the local telephone company, and this is in a world in which that telephone company faces competition from some of the people that are sharing its lines.

So my question is, is it even plausible to think that this phone company, or any of the phone companies, would be better service oriented if they didn't have to face any competition at all in their networks?

That was a loaded question, and we'll take the responses.

MR. HAZLETT: Now I know how you came up with the title for this thing. (Laughter)

The question is deregulation. You slipped competition and monopoly in as a surrogate, a proxy saying if we deregulate them and they have a total monopoly how are they going to be better? If you deregulate them today, or DSL say unbundling, they've got, all the ILECs together have about a 20, 26, 28 percent market share. I mean cable's dominant in that market according to DOJ standards.

The question is one of regulation and the incentives of regulation. And all through this and all the comments here that I hear about are the ILECs evil, are they incompetent, they're not doing anything with the technology. They're all exactly what you predict from disincentive effects for regulation. And particularly the modularity—Look, we can match DSL stories. I can give you ISDN stories, so I'll see your DSL stories and raise you ISDN. All this regulated platform stuff invites exactly the kind of fingerpointing and the non-ownership of service problems that is—

Look, how could you lose to the cable companies? I thought you were going to point this out. They make movies about cable guys they're so feared in America and the service quality is so bad, and DSL is losing to them.

Now you can sit around forever and talk about how evil and incompetent and how the world is being destroyed by these people lobbying. This is an outcome of the regulated environment that these companies are in. The question is how to improve the incentives in the marketplace to get better service.

I think that's where you have to come out in terms of taking all of these empirical observations.

If I can say one thing about the cannibalization issue, and I do have a paper on the AEI-Brookings Joint Center for Regulatory Studies web site on this, that it talks a lot about the cannibalization issue in cable.

The way you state it is only 50 percent incorrect. There is a cannibalization issue. The cable companies make a lot of money off of plain old cable service, video. And if they gave you high speed, real high speed, not what they give you today, but if they gave you real high speed access to the internet at the same price they could take away some of the demand for their current service. There's no reason an unregulated cable company, and they're unregulated, has to worry about that. They can always price for the extra quality they give you. They don't have to sell you the 20 megabytes per second product in place of the 15 which they're giving you today—those are maximum numbers of course, at $40 a month. They can price more. In fact the [DOXIS II] standards for the cable modems actually calibrate and can price how much bandwidth you're getting and the cable industry can charge for that in an unregulated environment.

But there is a problem with cannibalization and it's not the pricing problem that you suggest. Why sell for $40 what you can today sell for $80. Meaning today you can get $40 for cable and $40 for high-speed access. If you just gave people high-speed access that wiped out the cable revenues, why would you do that?

Well, very simply you could just charge them $80 for a much higher bandwidth product. That's the answer to that. But they don't do that. In fact look at the cable architecture. You've got 750 megahertz on the state of the art system today and how much do they use for high-speed access? Six. One video channel. How much do they use for digital cable? Well, 150 to 200 for digital cable. What's digital cable and what's subscription video on demand or plain or video on demand? That's internet TV. But they use so much more of their band, they give analog TV about 400 or 500 megahertz.

The reason that they use all of this bandwidth for the old services and the new services that are regulated the old way is because they fear regulatory appropriations, they see you out there, Charles, they've read your book, and they know that you are a structural separationist. You are an advocator of common carriage. And when you look at the video product you see access to the internet as a product that ought to be unbundled. And certainly it's not just you but it's the entire regulatory movement originally led by GTE and the phone companies attacking a competitive rival for local access. That's where some of this phone company lobbying money has gone.

It's very interesting that they want to regulate their direct competitor with common carriage rules. That's the AT&T position on structural separation today. You have to figure out why direct competitors want to make their rival more efficient. Obviously they don't. They want to screw up the investment in the rival sector and the rival delivery system.

But the fear of common carriage is what's keeping cable, they know that these open access rules are hanging out there and they've done a great job as an industry in fending this stuff off and sort of quarantining the problem to a very small amount of their bandwidth, but they don't want to make a big run at this and literally open up a lot of their architecture to common carrier rules that will take their Q ratio down to where the phone companies are. They like being an unregulated telecommunications provider and they're going to stay there and if they get some open access rules and the meaningless things and the Time Warner/AOL merger and stuff like that, that's fine. They can handle that stuff and lease access and the occasional must-carry rule. They'll quarantine that stuff and limit it as severely as they can. But those types of common carrier rules, and we talk a lot about them in the paper. This has been a 50-year regulatory battle in cable, those kinds of common carriage rules are exactly what they do fear, and that's the cannibalization issue. Not the pricing issue which an unregulated firm has no fear at all of cannibalizing, they just price for the value they're creating.

MR. FERGUSON: Where to begin. So many opportunities, so little time.

First of all, the cable companies do, I assure you, have concern about cannibalization caused by open architectures, video.

With regard to the second cannibalization issue which Hazlett referred to, shocking as this may seem I am in partial, only partial, but in partial agreement with him in that it is unquestionably true that there is an issue to be faced about the incentive to make large investments in this area. So that's my partial agreement.

The solution to that is not to kind of do nothing, however. I think the solution to that is to put in place broad incentives for investment which would make investment life easier both for the cable companies and their prospective competitors. But the really important thing to say is that the cable companies are a side show now. The idea that the cable companies dominate broadband is a very cute, quaint, incorrect idea. They have a higher market share in residential broadband which in bandwidth terms and in revenue terms is currently much less important than business services. In business services they are not a factor at all. And T-1, for example this year is probably about a $15 billion maybe $20 billion business, growing about 40 percent per year. It's been growing about 40 percent per year over the last five years which is really quite striking because the internet during that time was doubling, tripling, quadrupling every year and the price performance of T-1 has been essentially flat.

So about the telephone companies. The really interesting question is, is their behavior caused by the current regulatory structure or is it caused by their monopoly status? And then if it's caused by their monopoly status is it because they're incompetent or because they're ruthless?

It seems comparatively unlikely that the majority of their behavior is caused by their regulated status. T-1 services, for example are not heavily regulated at all. Residential dial tone rates are regulated but they're increasingly regulated by price cap and not by direct rate regulation. The enhanced services where they make most of their money in which they tie very closely, strategically and technologically unnecessarily closely to dial tone, voice forwarding, caller ID, voice mail. Those are much less regulated, sometimes not regulated at all than basic dial tone service. Then again, look at comparables. Look at IBM. IBM was essentially not regulated, still got very fat, happy and incompetent. And then another interesting case, look at AT&T's divestiture, got much less regulated, had to face competition for the first time, did very poorly, was a total joke in all of digital information technology.

When you look at the universe of explanations you're forced into the conclusion that it's a very substantial fraction, in my view the overwhelming majority—not all, but the overwhelming majority of the behavior you see comes from overwhelming monopoly status and not from regulation. That's not to say that regulation is always a good thing.

I would say that my view is the ultimate deregulatory view. If you chop these guys up intelligently then you don't need regulation and it should and would go away.

Q: Lisa Sockets from George Mason Law School.

Mr. Hazlett, you suggested that the unbundling regulations should be lifted for new investment, and I'm wondering how you determine what are the new and what are the old when they're providing broadband, and whether in fact your argument would also go to the unbundling obligations even on the old, the current network.

Q: Bob Hahn

Tom, I wanted to know—Charles said he thought broadband could be a very big deal. I wanted to know your views on that.

And Charles, I was wondering if you could flesh out a little bit your proposal for structural separation. Is it just going to affect the phone companies? Would it affect the cable companies? Anyone else?

Q: Kelly Bowen, Western Telecom Incorporated. I do analysis of all of these markets for private sector clients.

There was a question about Canada. I'm not going to address all of Canada because I wasn't paid to study all of it, but I can tell you why broadband penetration in the Maritimes is much higher than it is around here. Back in 1997 Maritel and all of the associated components in the Maritime phone companies selected a DSL technology and went with it. It wasn't a good one, they knew it, but it was there. So they beat the cable guys out the door by a lot. And they'll tell you the technology they selected wasn't the best one that might develop, but it was deliverable then.

Which brings me to the point we should get to here I guess which is never mistake for mendacity that which can be equally explained by lack of technological comprehension.

The problem with DSL was it doesn't work. The technology, ADSL is based on video dial tone. Forty-six percent of the people in the United States cannot access it because they live behind digital subscriber line multiplexers. They won't pass the signal. Until they come up with a way to pass that signal around the line concentrator or the new technology they're coming up with, hybrid amps which have their own problems, 46 percent of the people can't see ADSL which was going to be the telecom salvation to all of this.

Cable people, I work for a cable company. I can assure you the one thing they don't want to do is become a public utility. In fact I was told once when I came up with a bright idea to deliver business service over cable that not only were we not going to do it, if I came up with that bright idea in public I would be terminated because they didn't want to be a public utility.

We all hear about 5-9 reliability. Anybody got 5-9 reliability on their cable system? Three four's maybe. Would you put your business on cable? I won't even put my home telephone on cable. And I'm one of those people that's being passed by a service it can offer them. I don't want to do that, it doesn't work.

The other thing that we're not looking at here is the issue that evolution is not a constant wave moving through the world. We can have 2.4 terabytes of broadband on a fiberoptic pipe but what do we do with it then? You can't break it down. It's sort of like having a 5/8 garden hose standing next to a 96 inch water main and saying I need to water my yard. No way to hook it up.

The solution, the technology standards are out there. The problem is cost-effective implementation of the switching isn't. So rather than sitting here worrying about policy decisions that we should be making on things, I think we should probably take a little longer look at the cultural imperatives of the organization.

I've spent my entire life fighting with the phone companies and I have a low opinion of their ability to innovate. On the other hand, I used to work in the cable industry and they're (inaudible) either. So the old Sims thing saying our best friend is an educated customer. In the case of the broadband industry it may be their worst enemy, but I think everybody needs to get a little more educated about what really is available, what really is deliverable and why you can't get what you want.

The other question I would have for the panel up there is what would the customers want to buy, assuming they had broadband to the house? Has anybody seen any competitively provided content that you'd be willing to spend $100 a month on?

Q: I'm Bob Rogers with the GIIC.

I just wonder in commenting on, somebody asked the question about Canada, when I look at these bar charts of per capita deployment rates, Korea really sticks out. What's going on in Korea?

MR. LITAN: Okay, we've got enough questions.

One short answer, by the way, to Korea is lots of tall buildings—easier to hook up. I'm being only half facetious, but it's true.

MR. HAZLETT: I've just recently taken a look at the Korean market. There is a lot of density there and DSL is going well. If it doesn't work, something's different about Korea. But DSL is the dominant market share there, much higher than cable. And in fact the unbundling requirements have not been a factor. You have vertical integration for DSL provision there. You also have some considerable amount of broadband subsidy in the infrastructure, although many would argue that if you actually counted up the U.S. subsidies, talking about some. But there's a lot of subsidy in the U.S. structure that's probably largely wasted in terms of some of the current universal service programs or E-Rate programs and things of this nature.

But Korea is an interesting example. They've recently put on some unbundling regulations. We'll see if they actually kick in to give competitors market share over a shared network. But they've actually amassed this huge international lead without unbundling so that's an interesting metric.

In terms of Bob's question. I agree with Charles that this is potentially a very very big thing. I think the killer ant, you never know what the future holds, but the killer ant-like consensus is video, and maybe interactive video and gaming of these things. But this is why competition with cable companies is so much a part of this, because the cable companies are doing very well.

People talk about broadband, and I like to think I've been studying broadband since 1981 because that's what the folks across the street here at the National Cable TV Association, excuse me, the National Cable and Telecommunications Association purport to have offered the country for the last few decades. But who knows once you have the platform out there and you have the applications being written, because the market is created to receive them and generate revenues for someone, who knows what will come out of this?

MR. FERGUSON: First of all, my apologies for not responding earlier to the question about the Canadian situation. Unfortunately my answer is I just don't know much about the Canadian situation.

The Korean situation is interesting. Population and building density is certainly part of it. Financial incentives are probably part of it too but I don't think those are probably the largest part of it. The largest part of it is that the government said to Korean industry, you're going to do this. And that's why penetration rates are so high.

Let me talk a bit about the ways that structural separation could, should, would work. One thing that I'm not particularly a fan of in the case of the telephone companies is just chopping off the cabling, the last mile cabling from everything else.

What to do—Here I must admit that what to do is really complicated. The telephone system is very very complicated. So there is what would be ideal and then there's what you might be able to achieve.

What would be ideal would be take a look at the personal computer industry and the structure of that industry. How does that industry work? The way that industry works is there are a couple of proprietary technologies—Intel's got one, Microsoft has one. You can argue that Hewlett Packard has one in printers. But the overwhelming majority of the architectures and standards that drive the industry are non-proprietary. USB buses, dot, dot, dot. There's a lot of them. Very large numbers in display standards, interface standards of various kinds. And you have this hyper-competitive industry which, however, has to continue making itself standardized and that standardization is enforced by Intel, by Microsoft and by the collective self-interests of the people in the industry.

In the case of telecommunications one can argue that the importance of standardization, compatibility, and interoperability are so large that you might have some kind of mandated government or official requirement for interoperability and standardization. That might slow things down somewhat. But still, the goal would be that you would have a thin layer of standards and architecture and you would have a Silicon Valley- like infrastructure, very competitive, a lot of entry and exit, very dynamic industry, a lot of inter-capital funded startups.

Now how would you get there? You could get there I think if you separated the central office from the enhanced services platforms from the last mile.

There is an interesting thing about that and it's another facet of the telephone companies' dislike for real broadband service, real open broadband service.

If there was very inexpensive, very high speed broadband service available kind of everywhere all the time, then the need to get access to the telephone company's central office in order to compete with the telephone company, and currently you do need to get access to the central office, that need would go away. If you could get from anywhere to anywhere and swap services and swap computing technologies and computing resources, the world would be very different. Anybody who wanted to could start offering voice mail and call forwarding and caller ID and all kinds of fancy enhanced services and they wouldn't have to be inside the central office. All they would need is enough bandwidth, if the bandwidth was inexpensive.

A final comment on the cable situation. In my first software company we had an engineer who was not doing his job. We fired him. The way we fired him was his team leader went into his office and said you know, it's okay if you're the least productive engineer in the group because somebody has to be. It's also okay if you're the first guy to leave the office in the evening because somebody has to be. But it's not okay if you're both. (Laughter) That's what you say to the cable industry.

It's okay if you own proprietary content and it's okay if you have a closed system architecture that prevents people from delivering bandwidth to compete with you using your facilities. It's not okay if you're both. Choose one. Then I think I'd be happy.

MR. LITAN: Any final questions?

MR. HAZLETT: have an answer to a couple of the questions. Is it okay to—

MR. LITAN: Yeah.

MR. HAZLETT: First of all there was this question about real options, is there any price. I think the answer is no, there's no single price of access that the regulators can come up with that's going to compensate because at a high price people just aren't going to show up unless things turn very favorable for the industry, and at a low price there will be too much utilization.

This is why contracts, long term contracts, vertical integration, these kinds of institutions take place in the market to solve these incentive incompatibilities and come up with some complex equity ownership arrangement, and single price access rules don't do it. I think that is the answer. Now if you've got another take on it, please let me know.

On the question of how do you identify what's new, there's a lot of money being spent to identify the regulatory rules, some of it would go into that unfortunately. It's not going to be without debate.

Within the architecture of the phone system, and we've heard a little bit about it now, there are some fairly standard, because it's both slow to change and highly regulated, there are some fairly standard architectural guidelines to go on and the phone companies, the LECs are making the argument that in fact extending fiber out from the central offices past the three mile radius that DSL now effectively covers, according to some, ineffectively according to others, that those extensions of the fiber is essentially what has to be deregulated. Who knows? That would be unfortunately a regulatory process.

Can you make the argument that you should do away with the whole thing? Of course you can make that argument. It's a much stronger argument. Because you're not going to get any more old investment by doing away with access to the old regulated loop. It was built on the assumption that there would be access and in fact this is the architecture and the system that we've come up with. So we might have that debate at some point, but that's not today's debate.

MR. LITAN: If there are no further questions I want to have all of you join me in thanking our panelists for an excellent and stimulating discussion. Thank you.

(Applause)