George Gilder is chairman of Gilder Group Inc. and a senior fellow at Seattle’s Discovery Institute, where he directs the institute’s program on high technology and public policy. Author of the book Telecosm, Gilder showed up as a surprise guest at WTF 2004 and graciously agreed to take the stage when it became clear that the morning keynote speaker would arrive late. What follows is a partial transcript of his talk.
David Isenberg: In 2002, it came out that the United States was ranked No. 11 in per capita broadband penetration in the world. I took the growth rate from 2001 to 2002 and naively plugged it in from 2002 to 2003 and reranked everything, and look, the U.S. falls to No. 15 in 2003. Korea falls to No. 2. And if, for 2004, you naively plug in the same growth rate, look what happens: the U.S. is still No. 15. No. 1 is Finland. That’s no surprise. Switzerland is No. 2. And the surprise is Japan. But of course, I don’t want to be naﶥ and assume the same growth rate. George Gilder was a communications revolutionary when we were still in diapers. He’s also proved to me, as a hopeless liberal, that the communications revolution has no party. You have to find your allies where they exist.
George Gilder: I recently came back from Korea, and it’s important to really understand what happened over the last five years. A trip to Korea can give you an understanding. We didn’t have a fundamental bubble that consisted of Ponzi schemes and accounting frauds. That wasn’t the basic thing that happened. The basic thing that happened was that we launched a broadband revolution and didn’t consummate it because of regulatory mistakes. So it moved to Asia. Korea has 40 times the amount of bandwidth that we do. And they accomplished that in three years. When you have 40 times the amount of bandwidth, that’s 75% penetration.
When you have a true deployment of broadband in a country, including wireless broadband, the whole economy changes. In 2003, there was around $450 billion a year of commercial transactions on the Internet in Korea. A third of their economy was transacted on the Internet. If we had a third of our economy transacted on the Internet, versus our 2%, the business models that were deemed Quixotic and absurd because of the bubble would have succeeded. Grocery transactions are done on the Web all over Korea. Webvan and those other dotcoms would have prevailed there.
What happened has still not been grasped at all. The Internet movement that we launched here, that we invented, that we financed has been taken over by Asia. All because of an absurd regulatory situation. Telecom is the most thoroughly regulated and highest-taxed industry in the U.S. just short of tobacco and alcohol. Telecom is not a commodity. The chief insight that emerges from Clay Christensen’s disruptive innovation is that integration makes sense when you’re in an undershoot industry. Modularization makes sense when you’re in an overshoot industry. When your industry is in drastic undershoot of the potential of a technology, you need fully integrated systems. You need companies like Corbis, which does the only all-optical network. Every interface is Corbis technology. That reproduces the original AT&T model. It’s a broadwing system, but it’s an integrated network. That means it can steadily advance and outperform the other networks.
The reason telecom looks like such an unpromising business to people today is because of all the regulatory bodies and jurisdictions, as well as the general assumption by politicians that they can run information technology. That has been the problem. We had an aborted deregulation, which was transformed into a massive reregulation that inflicted a paralysis. As a result, thousands of companies in Asia are operating at a terrific pace, reaping substantial profits.
Question: One of the arguments that’s always put out is that if the market wants it, the market will have it. How have they done that in Korea, if it’s not the market?
Gilder: The reason the market couldn’t happen here was because it was stifled by regulation. That very period, the entire telecom industry had to address a complete transformation of the amount of traffic the system had to carry. The moment they had to mobilize to incur huge debt to support that deployment, we had severe deflation. The chief effect of a deflation is to punish debtors. That hit all the companies that had incurred debt to deliver this new fiber infrastructure. That caused something like 1,000 bankruptcies. It was a systemic problem rather than an industry problem.
The other thing was the regulatory snarl that I described. Regulators tried to address everything in a local loop, and that’s impossible with a technology that’s risky and ever changing. Korea was enthusiastic about broadband. Their government was enthusiastic about broadband. They had industrial policies that were in favor of broadband. It didn’t have to be very strong. The difference was that they didn’t have an anti-industrial policy and a punishing policy hostile to communications. It is the most heavily taxed industry other than tobacco and alcohol. It’s the most diversely and perversely regulated industry in the country. That led to an ultimate paralysis, which is why the market didn’t show the demand for video teleconferencing and multiplayer video games. All government services are provided on the Net in Korea.
Isenberg: What do you think of the interstate highway system as a way of delivering roads?
Gilder: I think it does a fine job. The interstate highway system isn’t as dynamic a changing industry as telecom. The system that succeeds in delivering roads fails at delivering connectivity at the last mile.
Isenberg: What if it just delivered dumb, unlit glass?
Gilder: I think the fiber is being delivered. I don’t think the government should fixate on a specific technology. This technology is still advancing very rapidly. The all-optical network I mentioned will be more flexible and robust than the hybrid networks that are currently being deployed. The defense department made its choice for a new national optical network being built by the Pentagon, and they’ve adopted a retarded set of technologies.
Melissa Davis: If you turned all of the glass on, it would just be light. We have particular issues in this country which are not extant in Korea. If you sat on a national LAN, you could watch everything collapse because of the traffic between the east coast and west coast. We could light a galactic supernova in the glass that we have. How do you deal with the intermediary stuff?
Gilder: How does Korea handle heavy traffic peak loads? They have 4:1 deployment of DSL versus the 100:1 deployment we have here. They have several redundant networks over Korea. They throw bandwidth at the problem. They have overprovisioned last-mile connections. DSLANS are set up at four lines to each connection.
Gordon Cook: Supposing we got rid of all of the FCC, how do we get a parallel with what happened with NTT in Japan? NTT unbundled and tore their network wide open. If we disband FCC, can we expect that local carriers would open their networks? I don’t think so.
Gilder: I don’t think they should. The copper cage is their problem. The biggest business error I’ve seen was the RBOCS failing to deploy fiber to the home in the last 10 years, when they could have done it. Japan’s system worked better than ours does because they don’t divide telecom into 150 different jurisdictions. They could force NTT to unbundle, but they have a lot of other kinds of competition emerging. The unbundling of NTT wasn’t the critical factor. And the forced unbundling of local loops in the US won’t solve the problem.
Roxane Googin: This network is a wonderful tool that we have given the world. When you were in Korea, what did you see them use it for that was particularly effective?
Gilder: All kinds of government services go over the Net. Retailing in general occurs on the Net. What we see in Amazon is just the general way retail all over Korea is accomplished. It comprises a third of their GDP. They’ve got EVDO deployed all over the place, so you see people in the buses watching TV on their cell phones. There’s widespread delivery of television services over the cell phones. Even Qualcomm feels intimidated by the broadcasters in the United States. They fear they’ll be regulated as a broadcast carrier. I don’t think that their corporate business-to-business services are more developed than ours. They’re probably less developed than ours. But they do do a lot more video teleconferencing and video messaging. Their broadband deployment exceeds their use of it. The measures of traffic I could find do not indicate the vast increase of traffic that could be enabled by this vast increase in traffic. The way they deal with spikes is that they’re way overprovisioned all over the place. People are learning how to exploit this, but it’s really only a three-year event.