Building Japan's Information Infrastructure

Create new user demands through innovation
that brings 'increasing return'

Original Japanese article appeared in Nihon Keizai Shimbun April 16, 1993.


The Japanese government recently announced a new economy stimulus package which includes "New Social Infrastructure" program that will support the installation of supercomputers in universities, personal computers in schools and optical fiber networks to the home. But the proposed budget package based on construction bonds and raising local telephone tariffs are alone not enough or appropriate.

In U.S. they have years of accumulated experience and expertise of the Internet. Internet usage has experienced exponential growth rates recently, and aggressive use by the new Clinton administration's staff represents the tip of the iceberg in terms of wider use of the "Net". The new administration's emphasis on information infrastructure as part of economic and technology policy should be interpreted as the reflection of this explosion.

We in Japan should reposition ISDN, lower tariffs on leased line services, which are at least five times higher in Japan compared with U.S., and build telephone-oriented networks and computer-oriented networks in fundamentally different ways. It is critically important to create new demand and new markets for the networking business that will leverage 'increasing returns' from technological innovation. As an alternative to raising money by national treasury bonds or raising telephone tariffs, a new 'Digital Network Bond' is proposed.

* * *

Japan's "New Social Infrastructure" scheme "free-rides" on the information policy of Clinton administration Just recently, during February to March, as a part of new budget program to stimulate Japan's sluggish economy, a "New Social Infrastructure" initiative in the telecommunications and information fields was proposed by a number of leading politicians and MITI and MPT officials. On April 13, the Japanese Government officially announced an additional budget package deal of some 13 trillion yen which includes this "New Social Infrastructure" initiative, but in a rather vague, very Japanese fashion that adds up to a little less than 10 % of the total.

This sudden proposal of "New Social Infrastructure" is aimed to financially support computer installations in universities and research institutions as well as schools, and the construction of optical fiber networks (ultimately to the home), taking advantage of the sluggish economy that requires any kind of government initiative to be stimulated. As Mr. Hayashi, President of NTT (Nippon Telegraph and Telephone) U.S., clearly articulated in his article in the Nihon Keizai Shimbun (Japanese equivalent of The Wall Street Journal) on March 10, the new U.S. administration's information policy is spurred at least in part by Vice President Gore's "Information Superhighway" proposals that he pursued a then Senator Gore for the last couple of years. Japan's government officials and LDP party leaders look like they are trying to 'free-ride' this external factor to introduce a new policy set in Japan. It is an example of the "Gaiatsu" technique often used to 'change' conventional structures in Japan by introducing foreign threat or pressure to help suppress any opposition to the change.

One of the focuses of this 'New Social Infrastructure' is the "Fiber-to-the-home (FTTH)" project. Under the new budget proposal, MPT, under their own initiative, is planning to start a 'pilot project' as a test bed to the nationwide FTTH. In my observation, MPT intentionally 'misunderstood" that during the Presidential election campaign the Clinton/Gore team committed any U.S. government under their administration to providing FTTH, and therefore MPT required that Japan should also start a similar project in order not to be left behind. As I understand from the other side of the Pacific, the new administration has not committed to the government laying FTTH, they are proposing high-speed computer networks to link research and education communities under NREN and NII (National Information Infrastructure) initiatives. It looks like rather than having government money pouring directly into FTTH, even if they had such money available, the new administration is planning to stimulate private sector investment by the telephone companies as well as cable TV and computer companies. There are also strong opinions that new digital wireless technology will require far less money than laying optical fiber.

In March, Japan's MPT submitted a proposal for FTTH policy to be discussed by the "Telecommunication Council", an official committee composed of selected industry representatives and scholars as well as other opinion leaders. MPT's underlying thinking is that "new information infrastructure should be built by the government, not by NTT which is suffering from decreasing profit". That means that MPT wants to reposition its leadership in the telecommunication field, not losing its control as happened after the deregulation in 1985. NTT on the other hand was reported as saying that they counter-argued against MPT by claiming that "NTT can invest in FTTH by itself, yet to do so we first need to 're-balance' the telephone tariff structure, i.e. raising unprofitable local charges".

The Ministry of Finance (MOF) joined the debate in support of NTT. MOF's support of NTT's position was in part due to the hope that it would help 'raise' the once hopeless NTT stock price. MOF may have been successful, NTT stocks gained by more than 60% within less than a month of the announcement, and this rapid rise in NTT stock played an important role in activating the entire Japanese stock market. For MOF, the scheduled opening sale of JR (Japan Railways) East's shares --which as a nationalized company is "owned" by MOF-- was the foremost concern and unless the Japanese stock market improved their fate is gloomy. Who should bear the cost of information network of the future? Buyers of national bond or taxpayers or users of local telephone networks? None of them seemed to be enough or proper.

Reposition ISDN first, before talking about "FTTH"
Let's look at the down to earth reality. For years, the main player in the next generation networks was considered likely to be ISDN (Integrated Service Digital Network). Where is it now? In Japan, commercial ISDN started in 1984, one of the first in the world. Despite vast investment already made, usage of ISDN in Japan is not as promising as once expected. ISDN was positioned as an alternative to existing analog telephony, yet in actuality uses have been centered around computer-to-computer communication. It's a matter of fundamental concepts and understanding of the service. A very important issue by itself.

First of all, we should reposition ISDN. Whether we will position ISDN as part of computer digital network service or digital alternative to existing telephony should be the number one question. My opinion is, at least for the time being, the former. Even though they say it's the same 'digital', telephone and computer networks require different optimum economic set of solutions and a too simple integration/mixture is questionable. The distribution of demand or traffic on telephone and computer network cannot be one and the same.

In Japan, there have been too many 'regional informatization programs' and pilot projects of new technology and services, they have tended to create much confusion rather than knowledge. Although quite a few 'third sector' businesses were established to provide CAPTAIN (Japanese version of videotext service) and CATV, few if any have balance sheets that meet original expectations.

The Hi-Vision (Japan's HDTV) project initiated by NHK and supported by MPT and MITI was late to catch on to digitalization and their proposal submitted to U.S. FCC was finally turned down this March. Direct satellite broadcasting and satellite communications businesses are both suffering and facing a very critical moment. Their current demand forecast is far less than original hopes and there is little expectation that capital investment will be paid off in near or far future.

Why are there so many failures? One answer is that the relationship between 'contents' and economy of services was not considered, that is, to objectively analyze the simple demand and supply dynamics of rather 'unknown' kind. There has been one successful case. PC communications service, such as NIFTY Serve or PC-VAN, similar to Compuserve or Prodigy in U.S., achieved significant growth despite low expectations and predictions.

For the Japanese, a humble, open and objective retrospect is essential. If not we may follow the trouble of JR, which constructed too many local tracks where economically effective demand was non-existent and defeated by the increase of automobiles and the road.

Understanding the dynamics of the Internet is the key
Even the White House is now using it!

The situation in U.S. provides a good contrast to that in Japan. There is massive amount of experience using computer networks in U.S. Learning from the history and current situation of the Internet makes a big difference in understanding the information policy strategy of the new administration.

Even in Japan we can now obtain raw and rich information related to the information policy of new Clinton administration via the Internet. For example, when Mr. Clinton announced his economic policy in February from a 'town meeting' in Silicon Valley, only a couple of hours after the meeting finished we received a full transcript of the President's and Vice President's statements and the question and answer session the participants at Silicon Graphics via Internet. We know that the new administration's staff are using the Internet and other networks extensively, usage that must have enabled them to 'feel' the power and the potential of the Net, thus making full effort to implement the information policy. As far as I know, our government and political party leaders are far behind the rest of the world in their appreciation and understanding of the importance of these issues.

Now, for the Japanese to learn, let us look back at the history of the Internet. The Internet has its roots in ARPANET started in early 1970s to link research organizations working on military contracts. In the 1980s CSNET was built with initial funding from National Science Foundation, then NSFNET grew from these earlier efforts in order to, as officially stated, link supercomputer centers nationwide , and became the core of the vast 'network of networks' --the Internet-- that now interlinks research centers, private companies, libraries and schools. All of this was achieved in an atmosphere of open and voluntary collaboration of users interacting with each other, exchanging information and sharing resources and results of research activities. This is the Internet of today. And the new administration's information policy is being built around the Internet or at least its example.

The Internet has increased its users exponentially over the past few years. Now there are more than 120 countries interconnected with estimates of more than a million host computers and more than 10 million users. It is by far the largest computer network on earth. It used to be rather science oriented, or computer engineering dominated, but recently the coverage and contents of the services as well as user communities underwent a significant change. Use by the humanities and social science communities, education, private business and ordinary citizens is increasing by leaps and bounds. This trend is not only to be seen in the advanced countries of the North, developing countries of the South, from Africa to Asia as well as former Eastern socialist countries are all entering into the Net community.

Where to begin for the Japanese?
First, lower the price of the leased line services

In this massive explosion of the Internet, Japan was left behind. Almost 10 years after the U.S., a Japanese version of Internet was started to link universities and the research centers of national and private companies. The core of this effort was mostly the result of a private, voluntary research program named WIDE (Widely Distributed Environment) lead by Japan's networking guru Jun Murai of Keio University. WIDE has its own leased line trunk, but the speed of this 'backbone' is only 192K bit per second, almost 1/200th of American backbone of 45Mega bit per second, and the capacity is far less than that.

Why is that? The biggest factor is the high cost of leased line services in Japan. The Internet developed by linking computers directly via leased line services. However, in Japan, leased line services are prohibitively expensive and thus offer a bottleneck for the computer to computer networking. One example shows that the price of a 1.5 Megabit line costs 5 time or more in Japan than a similar line in the U.S. (see table). There may be some differences in market competition or quality of the lines, as Japanese service providers try to defend, but this difference is far from understandable.

Leased line services in Japan is making huge profits. If we take NTT as an example, the leased line division accounted for only 7 % of total sales revenue last year, but contribute to some 15% of the total profit of NTT's entire telco business. The gross profit as well as profit ratio of leased line services are second to those of the long distance telephone service. NTT is asking for a 'rebalancing' of the telephone price structure by raising the local telephone tariffs, but if that is allowed then the highly profitable leased line services should also be significantly rebalanced.

By doing so, many regional universities, research institutions and small private companies could also utilize high speed network circuits and in turn Internetworking in Japan could also enjoy rapid expansion. This should be the information infrastructure today. NTT and other carriers could expect a vast new demand for their services, this would spur R&D efforts and innovation and I believe lead to significant increases in revenue and overall profit from leased line services at the same time. It would be a healthy win-win game for sure.

Innovation that leads 'increasing return' will boost dramatic lowering of communication costs
That is not enough, however. It is of critical importance to learn from semiconductor innovations and the recent price revolution of PCs, that was based in part on those chip innovations. Technological innovations made possible the 3 to 5 times increase in capacity and performance of PCs while also allowing for significant reductions in prices, thus creating new products, new services, new demand and new industries in a 'chain-reaction' manner. Consumers of course welcome this kind of keen competition based on "Increasing return and decreasing cost" that we've seen in the PC industry, a new economic principle expanding the existing paradigm of conventional economic theories.

That is so far what happened to the chip industry and micro electronics industry including computers and consumer electronics. But not as far as I know to the telecommunications industry, at least in Japan.

The Network Industry, both telephone and computer networks, should learn and realize the same lesson. By the innovation and evolution of chips, inevitable 'downsizing' of switching machines and routers or other communication equipment, breakthroughs in light wave communication technology, etc., the 'density' of networking will exponentially increase. It will result in a sharp drop in transmission costs, switch or any other type of carriages of electronic signals in mostly digital format.

We may not need many trillions of yen to invest in the new infrastructure. So called 'information appliances' and digital mobile telephones (Personal Communications Systems) will accelerate this trend. Just as happened to the CPUs and memory in the chip industry, all sorts of telecommunication or networking prices may be easily dropped by new comers, that lead the creation of brand new demand not existed before, pay off the investment within reasonable timeframe for further expansion.

We can expect the emergence of 'network publishing' that will provide revolutionary low prices for information all in a digital format. The price of information today mostly does not reflect the cost of creating the contents, which should be the real value, but consits of the costs of storage and efficiency of delivery of information, i.e. in essence the cost of paper, printing, transportation, warehousing and other physical distribution mechanisms plus personnel costs attached to these 'extra' services. But if the Internet is to be the delivery mechanism and compuiters used for storage, then the cost of constructing electronic networks will allow net reduction of delivery cost by orders of magnitudes while giving ample volume of paths to reach information the way users want direct. The real significance of the explosion of Internet is that it demonstrates the ability to allow anybody to tap into the vast supply of information in a very very inexpensive way. That is what our government and industry people in Japan should learn from U.S.

Put the priorities to the users, invest the future with new kind of 'Digital Network Bond'
We must come to think of users first, how and what kind of intellectual infrastructure for citizens can be built, before we begin laying the physical fibers. Lowering the cost of leased line services today would facilitate a 'network of networks' to develop openly in the manner of the Internet expereince in the U.S. We must do everything possible to lead us in this direction both in prices and technological or even social and political environments.

In doing so, as in the U.S., new experimentation that may dissolve the barriers between broadcasting and telecommunication will be tried with users in mind, new networking technologies will be created one after another, new demand will be created by lowering price and raising performance. These new movements will reveal, in more natural ways, what role the government should and should not take and what role NTT and other communication vendors should play.

As for the resources of new information infrastructures, I would like to propose the introduction of new "Digital Network Bond". In Japan, after the near total destruction of telephone network during the World War II, 56 million subscriber lines were built in just 30 years. How? "Telephone Bonds" were sold to the new customers and these bonds financed a large portion of the costs of building the telephone network.

A Digital Network Bond would allow direct borrowing from the new users of the new services, not indirect subsidization from local telephone users. We must note that these user clusters are awfully different, at least during the long early stages. If a long term bond were issuedthen it would be a far more stable investment than speculating the stocks and would pay-off in the long run. Nobody will lose the game.

In Japan, public investment is often chanelled through special interest groups of limited corporations or industries, along with those of some ministries or politicians. These are the sources of most of the recent political scandals. We do not need this kind of investment in the communications business. If we want to really learn from U.S., we must learn much more deeply instead of making free-ride choices on the surface. Our ministries should, like the White House, use the open network to provide a direct information exchange with the people, make available information that bureaucrats and large corporations tend to monopolize, create open and frank debate by all parties concerned before the eyes of the public.

It is also important to promote more geographically distributed network design. Do not centralize the power around Tokyo, but provide a foundation in regional economies. Realizing an autonomous distributed network society, which is the real essence of the Internet, will be the most critical issue for the success of the information and communication revolution of the coming century or millennium.