W. Kenneth Ferree, Chief

Cable Services Bureau

Federal Communications Commission

at the

Broadband Outlook 2002 Conference

"How Do You Build The Information Superhighway?"

Remarks As Prepared For Delivery


Good afternoon. I want to thank Dick Wiley for that kind introduction and thank the sponsors for inviting me. Iím pleased to be here with you today for Broadband Outlook 2002.

While I was preparing my remarks for this conference, I came across a report that former President Clinton and Jack Welch each receive about $150,000 per speaking engagement. I thought to myself, what in the world could either of these fellows say that would be worth that much money?

Now Iím sure their presentations are polished and professional, perhaps even entertaining. They probably tell a few jokes, maybe relate an anecdote or two that would give one a glimpse into their respective worlds, and then, no doubt, offer some of their well-developed, but probably not new, views on current issues.

I may be cheap, but Iím not sure Iíd pay $150K for that and a plate of chicken.

It does offer me a sort of disclaimer, however. Because, while you will also likely not gain any great insights from my modest presentation today, you will have spent no where near $150,000 bringing me here.

So Iím really a very economical choice for a speaker.

Iíve thought of trying to make a living at it but, wellÖ. You see the problem.

Which leads me to the substance of my remarks ó economic choices by consumers and the development of broadband infrastructure and services in the United States.

Now youíve heard from a number of sources that the broadband issues being debated today at this conference are central to communication policy in the first decade of the 21st Century. Indeed, Iíve seen quotes from people in industry, consumer groups, and government ó from both sides of the political spectrum ó suggesting that broadband deployment and development will be the key ingredient in a robust economic recovery.

One study suggests that widespread broadband deployment may result in an additional $500 billion a year of economic output by the U.S economy.


So how do we get there from here?

Now you are all looking at me and awaiting the answer.

Allow me to remind you of my disclaimer ó I donít have any expensive pearls of wisdom to share even though, as Barbara Esbin, one of my Associate Bureau Chiefs has reminded me, we spend a lot of time chewing sand in this area.

And based on the importance that I see being attributed to the broadband revolution, it might not be a bad thing not to be the lead revolutionary.

Leaders of revolutions sometimes end up sacrificing themselves in the process of achieving their goals.

And if they donít, their followers may do it for them.

There are a number of historical examples of this happening, but the first one that came to mind was the "Maid of Orleans" ó Joan of Arc, who apparently was too dangerous to King Charles VII after she saved his crown, and other parts, from the English, but ultimately was burned at the stake.

My favorite example, however, is related by Jacob Burckhardt in his book, the Civilization of the Renaissance in Italy. There he tells of the citizens of a small town in Italy in the 15th Century who, failing to think of an adequate recompense for a soldier who had almost single-handedly saved them from invaders, decided that the only worthy reward was to kill him and make him their patron saint.

These stories give new meaning to the phrase: "achieving oneís end."

So pardon me if, for the time being, I settle for the more pedestrian role of observer to the broadband revolution.

In that role, I can tell you a few things about some of the grains of sand that we process as we look for wisdom.

Grain of sand No. 1: The suggestion has been made that if South Korea can do it, we can do itÖ and we can do it the same way they did.

Now it is true that close to 60% of all net connections in South Korea are what we would regard as "broadband" connections. That translates into 9.2 broadband connections per 100 inhabitants. By comparison, the rate is 2.5 in the U.S.

But ó this will come as a shock óthis is not South Korea.

The United States has a population of nearly 300 million people in an area of more than 3.5 million square miles.

The Republic of Korea, by contrast, has a population of about 50 million people, all within an area of about 38 thousand square miles. That is, it is about the equivalent of packing the populations of California and Pennsylvania in an area 3/4ths the size of Virginia.

Moreover, because of the rugged terrain throughout large sections of South Korea, most of the population is packed densely within a few large cities.

I think we could all agree that if this were just a matter of getting broadband to a few of the largest metropolitan areas in America, we could do that ó indeed, weíve largely done that.

But here, unlike South Korea, we are faced with a much larger task. Our challenge is to create a broadband environment for a large nation with a significant rural population and many smaller population centers. South Korea offers no easy answer for meeting that challenge.

Grain of sand No. 2: It has been said that open access to all networks is the only way to achieve a viable broadband market. We hear this one repeated regularly, but I have reservations.

Now certainly, "openness" in general is a good thing. But you canít stop your analysis there. What would be the specific public interest benefits of a pure open access model and what would be the probable costs?

As the ILECs are wont to point out, itís hard to think of a greater disincentive to network investment than a requirement that networks must be made available to all competitors at some regulated wholesale rate.

I just spent a couple of days at a conference in which there was a heavy ILEC presence. I canít tell you how many times I heard the ILEC folks talking about "privatizing costs and socializing the benefits." But they have a point.

Indeed, itís not entirely clear what the advantage of building a network is if it always is possible to use someone elseís. And, in a reverse chicken-and-egg problem, the "someone else" that you are expecting to build the network might never come along because of the lack of an economic return on the investment in new facilities.

This point is made persuasively by the National Research Councilís Committee on Broadband Last Mile Technology in their paper titled: "Broadband; Bringing the Bits Home." There they explain that the network investments that will be required to achieve ubiquitous broadband capability must generate significant returns in the short term, or they simply will not be made. Pure open access, the Committee goes on to say, may deter that investment entirely because it will prevent the facilities-based provider from capturing all of the benefits of its new facilities.

So we have to be wary of accepting, without due and thorough consideration, apparently attractive propositions like "open access."

Grain of sand No. 3: Another theory is that infrastructure deployment must first be ubiquitous, or nearly so, before we fret too much over adoption rates. This is the "build it and they will come" approach to the broadband revolution. It also is a favorite among those who favor government subsidies for broadband deployment.

The problem with this one is that it just does not appear to be so as a factual matter. By all accounts, a large majority of homes passed by cable, something like 70%, now have cable modem service available and almost half of qualified DSL homes have DSL service available.

Nonetheless, few Americans are making the economic choice to subscribe to broadband services.

Subscriber rates for broadband connections are in the single digits. And I donít think it is any mystery as to why. The fact is that there are very few true broadband applications that would generate consumer demand today.

And no matter what anyone else says, it is the consumer who will drive this revolution. When consumers find that broadband services provide them with some new value proposition, some additional utility, or even some new status, subscriber rates will increase and the gold rush will be on.

So what are the "killer apps" that will lead us to the mother lode?

Ah, youíre looking for pearls of wisdom again, and Iím just here chewing sand.

Now there are, of course, far out applications that might be termed Broadband science fiction. The only limiting principle on such a flight of fancy is that the applications must be rooted in, and dependent upon, high bandwidth communications.

The problem with these, however, is that current generation "broadband" networks cannot support them.

When we say "broadband" today, we are thinking about networks capable of delivering several hundred kilobits up to a few megabits per second of data downstream and maybe a few hundred kilobits per second upstream. The broadband science fiction applications may require tens or hundreds of megabits per second in each direction.

So I donít want to get preoccupied with these for the time being.

But the next generation of broadband network ó one that presumably will be symmetrical, or nearly so, and capable of delivering perhaps ten megabits per second ó should be able to support killer apps, the predecessors of which are staring us right in the face.

Napster and other Napster-like services, for instance, are good indicators that streaming media holds great promise as a killer app. Itís not hard to imagine, therefore, that a future broadband world may include a number of services providing not only music, but high-quality video services as well.

But this leads toÖ

Grain of Sand No. 4: There is the criticism that the current broadband infrastructure is technically incapable of providing an acceptable quality of streaming video to consumers ó and intentionally so say the cable critics.

The argument has been made that cable operators, by design, devote an artificially small amount of capacity within their systems to internet access in order to forestall the development of streaming sites that might provide competition to the core cable video product, particularly high-margin premium channels.

Assuming for the sake of argument that is true, what policy course does that suggest?

One might say, itís easy. If cable companies are deliberately choking off broadband capacity to inhibit the development of competing services, the FCC should forbid them from doing that.

As is often the case, every problem has an obvious and simple answerÖ that is wrong. In this case, there is no way that the proposed solution could be done surgically. That is, once the government gets involved in bandwidth management decisions, there is no clear exit strategy. Not only do we have a mouthful of sand to contemplate, we could find ourselves sinking into regulatory quicksand.

How much bandwidth is enough? Does it depend on the way the network is engineered? Does it depend on the identity of the operator? Does it change as compression technologies evolve? Does it change as other broadband networks evolve?

And that is not even getting into the legal and constitutional issues that such heavy handed government management would implicate.

No, the answer canít be that we in the government will dictate how broadband providers manage their networks.

Which leads me to Ö

Grain of sand No. 5: Why doesnít the government build the information superhighway? This question usually takes the form of some reference to the construction of the interstate highway system in the 1950s.

And for an answer we, again, look toÖ South Korea.

The South Korean government, after all, was heavily involved in the roll-out of broadband in that country. Not only did it set a date certain for widespread broadband deployment, but the state-owned, all-but monopoly, PTT, Korea Telecom, acted aggressively, and trained what the publication the "Economist" called "an army" of technicians to install DSL.

So why canít we do the same?

Remember the lesson from Grain of Sand No. 1, this is not South Korea.

In that country, the investment by the state-owned PTT in the project was within a manageable range because the vast majority of the population not only resides in a few large population centers, but also within multiple dwelling unit structures within those population centers.

Again, itís one thing to build an information superhighway when you are running from one MDU to the next, block by block, and sweeping within your service area a large chunk of the countryís population.

Itís quite another thing to build an information superhighway in a country in which cities are surrounded by large suburban areas with single family homes on relatively large pieces of property, where there are countless smaller towns and cities distributed over a 3.5 million square mile area, and where there are large agricultural areas that contain no central population center at all.

Those of us that live in or around a large city, and particularly here in the Eastern United States where you might be able to drive through five states in five hours, lose perspective on just how great the open spaces are that dominate most of the geographic area of the United States.

Iím always struck, when I travel by car west of the Mississippi, by the sheer vastness of the country and how widely dispersed the population is.

My father, for example, lives in a nice, but not opulent, home outside of Carson City, Nevada. His house sits on a 25-acre lot. All of the homes in that neighborhood do.

This makes for some very tired kids on Halloween night.

It also makes the suggestion that the government should take responsibility for, or subsidize, broadband deployment almost totally unthinkable, at least if the subsidy is to have any significant impact on the roll-out of the services.

And that is why the references to the construction of the interstate highway system are inappropriate. In that case, we were talking about building connections between hubs ó the backbone, if you will ó within a network of highways and smaller roads.

But the choke point in broadband today is not, apparently, in the backbone, but in the famous last mile. The proponents of the government subsidy solution really want an interstate highway system that goes to every personís door. I wonder how that would have gone-over in the Eisenhower Administration.

Of course, there is another, not so subtle reason that the construction of the interstate highway system provides a poor analogue for development of broadband ó the government actually runs the interstate highway system.

Now I have not heard anyone suggest that the communications networks should be nationalized and operated by the government, but one should ask whether ó and to what extent ó government would become involved in any such project. Because with government money usually comes some corresponding obligation on the part of the recipients.

And the balance between the government benefits conferred and the private obligations assumed is never very satisfactory. On the one hand, if the government demands very little, either in terms of public interest obligations or remuneration, it is vulnerable to the claim that the public is subsidizing someoneís private interests.

On the other hand, if the obligations are significant, it may be said that they inappropriately interfere with operations of the private market, and thereby limit efficiency.

And that criticism cuts pretty deeply. It cuts deeply because we have learned over time that private markets generally are much better at using resources efficiently, much better at fostering innovation, and much better at wealth maximization, than are command economies dictated by government policy.

So what is the answer? Well of course we look to South Korea for instruction!

Almost 75% of web users in South Korea access streaming media content on a regular basis. That dwarfs the U.S. rate of about 25%. And there are more than 800 web-casting companies in South Korea!

So what did they do right to make streaming media a much more widely available and acceptable service? Ironically, itís that all-American C-wordÖ

No, not consolidation or concentrationÖ


In South Korea, as a result of Korea Telecomís rapid DSL roll-out, there was an equally ferocious competitive response by cable networks.

The fierce competition between the two, and the presence of rival infrastructures, makes it difficult for either of the operators to protect their legacy services. If a DSL provider is offering access to 800 web-casting channels, you can bet the cable operator will as well. And as the DSL provider improves the quality of that streaming media delivery over time, the cable guy better keep pace.

Itís hard to worry about cannibalizing your own markets when a competitor, red in tooth and claw, is doing it for you.

And, itís not just the 800-channel web TV service that we are talking about here. When I think of streaming media I cannot help but think of one of the proven consumer drivers of all sorts of technologiesÖ Gaming.

With capacity for high quality video delivery, the potential for the creation of advanced on-line games can begin to be tapped. Now I confess that Iím not a big fan of most of what Iíve seen of computer games and software, but we would be kidding ourselves if we did not acknowledge the potential importance of this material for pushing broadband use and acceptance.

But that raisesÖ

Grain of Sand No. 6: So isnít, my mother asks me, all of this broadband hype really about gimmicks and games? Why in the world would I want broadband?"

She has a point. If broadband is ever to really take off, it will need to offer mass market benefits. Are there any hints on what that application might be?

I think there are. It may seem mundane, but broadband might thrive by making plain old telephone service better.

Again, Iíll use my favorite bad comparison ó South Korea!

It turns out that, because international rates are so high for circuit switched telephone service in South Korea, IP telephony products are a great driver of broadband adoption in that country.

But even where there is no strong economic driver for circumventing the existing telephone network, broadband offers the possibility of advanced features on the basic service that might prove to be significant.

Iím talking, of course, about a video-phone service.

At the end of the day, communications is about bringing distant people together. And thatís what telephony does.

Iím told the root of the word "telephony" is made up from the Greek words for "to speak afar."

There are, of course, countless applications for such far speaking. In the professional world, people spend a lot of their time, money, and effort getting to and from offices and meetings. Many of you probably traveled a great distance to be here today. Yet, the nature of professional services, which is trading in knowledge and expertise, rarely requires physical contact. Thus, one need not be Arthur C. Clarke to envision a high-end teleconferencing application that would obviate a great deal of professional travel. The ability to engage in sort of a "virtual visit" could be used for telemedicine, teleconferencing, teleworking, or any one of other similar applications.

Indeed, we already are seeing some early harbingers of this trend. There was an article in USA Today on the 22d of this month describing the way that "virtual interviews" are coming into vogue. How long until the dreaded Ĺ hour screening interview in some strange hotel room is a thing of the past?

And back to my motherÖ

Much as professionals like to "speak afar" with their business associates and clients, etc., so do the vast majority of people like to speak afar with their friends and family. If you want to sell broadband to my mother, you are more likely to do it by telling her that she can use it to have a face to face conversation with her granddaughter about clothes, hairstyles, and all that is hip.

To help illustrate the power of this point, I want to read to you a passage from the legislative history to the first information superhighway legislationÖ.

This is from a Senate Report from the 33d Congress, 2d Sess.Ö 1854.

The bill being considered involved a grant of some 2 million acres of public lands for a transcontinental telegraph terminating in San Francisco. In exchange, the federal government was to get 8000 words per month of priority use on the telegraph, free of charge.

Howís that for a bit rate?

Anyway, in detailing the benefits of the plan, the Senate noted the following:

But it is in its social bearing that the advantages of a telegraph to the Pacific will be most strikingly seen. Every hamlet, it might almost be said every home, in thirty states of the Union, has its representative on the Pacific shore. By the aid of a telegraph they would be in immediate communication with each other. Every message, whether of joy or sorrow, could be instantly transmitted either way; and sons and fathers, wives and mothers, whose relations are now a thousand miles asunder, would be, for the purposes of the interchange of intelligence, as it were, under the same roof.

That, if you ask me, is one of the great hopes of broadband. Families that now may be a thousand miles asunder, may, for the purposes of the interchange of intelligence, as it were, be under the same roof.

When people like my mother can use her connection to enhance her relationship with her granddaughter, broadband will cease being a niche service for high end professional users, and it will start to become a ubiquitous mass market consumer service.

At that point, regulators can all stop worrying about broadband deployment because the private financial markets will make sure that it happens.

Thank you for your attention and for coming out today. To get this virtuous cycle rolling, I am expecting that next year, at "Broadband Outlook 2003," they will offer the option of participation through videoconferencing.