Time to Push Japan's Telecom Policy Reforms

Shumpei KUMON (Executive Director, GLOCOM)

The U.S. Senate recently endorsed a resolution calling for a "Telecommunications Big Bang" in Japan. I agree with most of the findings stated in the Senate resolution--particularly the criticism of the serious delay in the spread of Japan's broadband Internet service and e-commerce, as well as the Senate's concerns of the effect that this delay will have on the Asian and global economies. I also strongly agree with the recommendations presented in the latter part of the resolution that "the government of Japan should take credible steps to ensure that competitive carriers have reasonable, cost-based, and nondiscriminatory access to the rights of way, facilities, and services controlled by NTT, the NTT Group, other utilities, and the Japanese government." I think there is still room for debate, however, over whether it is enough to seek the causes primarily in the telecom-related laws, regulations, and monopolistic practices in Japan. I suspect that some serious social and economic factors have until recently been causing this delay in the spread of the use of information technology in Japan.

Specifically, I doubt the effectiveness of the important first measure proposed by the Senate, namely, the request to reduce substantially the interconnection charges for telephony. I argue that the interconnection charge issue is not critical. I believe the goal of telecommunications policy should be oriented toward encouraging the fastest and widest possible uptake of the Internet and IP based networks, and not be dominated by arguments that are firmly rooted in the era of voice telephony. The most pressing need for Japanese consumers and industry today is not reduced telephone tolls. The most pressing need, as accurately pointed out in the Senate resolution, is the immediate acquisition of high-speed Internet access at a reasonable price, and the swift inauguration of all types of community networking activities and electronic commerce based on this service.

The rate of Internet diffusion in Japan is depressingly low, and remedying this situation should be our central concern. One of the primary reasons for the Internet's steady growth in the U.S. in the 90s was the flat rate system for fixed telephone applied to residential customers. Thus the main part of the Internet in the U.S. first developed as a virtual network over existing telephone lines. If Japan were to take a lesson from this, one definite requirement would be the swift introduction of some form of flat rate for Internet connection services. Although there recently have been proposals to introduce flat rates for some data services, it is too little, too slow.

However there are some notable successes in the Japanese telecommunications sector, particularly the rapid rate of cellular telephone diffusion in Japan, which is higher than that in the U.S. This growth is possible because in addition to corporations, the general public--particularly many young people--is using cellular phones for everyday communications. Short messaging services are very popular with young people, and information transfer, banking and simple web services such as "i-mode" are popular with older generations. Even though cellular telephone charges are higher than those for fixed telephones, cellular subscriptions continue to increase rapidly. In fact, over 53 million persons of a total Japanese population of 126 million own a mobile phone. There is intense competition among telephone service providers in this sector.

In contrast, the number of fixed telephone subscribers in Japan is steadily declining. Regardless of the charges for telephone connectivity, it is inconceivable that new players will appear in the market for both local and long-distance fixed voice telephone services. This has been made abundantly clear by the experience in the U.S. In the future, the mainstream of telephony certainly will take the form of IP telephony, offered as an integrated part of Internet services.

High-Speed Internet Access is Key
I suggest that the current discussion of interconnection and access in Japan is misguidedly focused on voice telephony rather than on DSL and IP telephony. The legal and policy framework for the deployment of DSL in Japan is, on paper at least, quite open. Definitely, there is strong demand for line sharing among ISPs. If discussions about telecommunications reform were to focus on ensuring that actual access to NTT's copper loops was as open as it is on paper, we should be able to achieve a situation in which access to high bandwidth data connections and IP telephony makes arguments about traditional interconnection costs irrelevant.

In a speech just before his resignation, former FCC Chairman Reed Hundt discarded the notion of a "five lane information highway". Instead he boldly argued for the necessity to construct a dedicated network for the Internet itself. Considering subsequent activity in the U.S., however, it does not appear that former Chairman Hundt's vision was fully realized. Indeed, while there has been rapid construction of optical fiber backbone networks since 1997, these networks seem to have been dragged back and subsumed by existing telecommunications models rather than form a new all-IP Internet-oriented infrastructure. The question of how best to provide high-speed Internet access remains.

Since the beginning of 1999, providing broadband access for Internet services has been recognized as a critical U.S. policy initiative. Although such a far-sighted policy is cause for celebration in itself, to achieve it the U.S. has relied primarily on DSL utilizing existing telephone lines and cable modem technology using unmodified coaxial cable. Since the enactment of the Telecommunications Act of 1996, full-scale competition has only recently begun in local service between telephone companies and cable companies, which had been regarded as belonging to different industries. However, this competition is for Internet access services, not for telephony (or cable) services. The FCC has applauded these developments and is continuing with a policy it describes as the "Unregulation of the Internet". Specifically, the principles underpinning this unregulation policy have been used to justify not placing harsh regulations on open access to cable.

I feel that the distinction drawn by the FCC between the unregulated Internet and the regulated telecommunications network is being blurred by these and other recent decisions. For example, the decision the FCC made last November on line sharing is not a telephony-oriented policy. Rather, it is specifically designed to affect the Internet directly.

Regarding unregulation, the promotion of competition and the principle of unlimited access to a company's network elements seem to be viewed as a trade-off. The latter must be abandoned to achieve the former. But, is this really the proper decision? Regarding Internet connections, if carriers want to preserve the value of existing network elements over the long-term, it seems more rational for them to unbundle their old network elements and allow them to be used by as many alternative carriers and ISPs as possible--at reasonable prices of course--even without regulatory intervention.

However, through merger and acquisition, incumbent and major telcos are bundling wired and wireless telephones, cable television, the Internet, and other services and offering them end-to-end. These end-to-end services are offered at a discount as a new type of service. It seems that American regulatory authorities already actively approve this development. Yet these major telcos do not appear to be attempting to combine these newly integrated business units on a unified IP Internet platform. Instead, they are trying to prolong the existing telephone and cable services, now generating large amounts of income, while offering new high-speed Internet access services as slowly as possible. It is natural that existing enterprises are reluctant to unbundle their network elements and sell them. But would it not be more proper for the policy-making authorities to keep the promotion of Internet use in mind and continue to demand unbundling and deregulation of existing network elements?

In the U.S. both research groups and government agree that the spread of high-speed Internet access is slow. As a result, dial-up connections through telephone lines will still account for two-thirds of connections even in 2003 and will still be at a 50% level in 2005. Of course, it is still possible to consider this spread surprisingly rapid. The number of subscribers to high-speed access is projected to rise by 50 times for DSL and 30 times for cable modems during this five-year period. However, during volatile periods in which time is measured in Internet years, I contend that this rate of increase is too slow.

The Perils of Complacency
In George Gilder's 1989 book, Microcosm, Gilder argues that as Japan became complacently self-satisfied in the latter half of the 80s and fell behind the cutting edge of the information revolution, the U.S. had already staked out a lead for the 1990s by innovations in microprocessors and telecommunications technology. He is right.

But what about the decade starting in 2000? Have the spread of the Internet and the explosive growth of e-commerce in the U.S. during the second half of the 90s created complacent self-satisfaction in the U.S.? When compared to the bold plans to encourage the deployment of new means of high-bandwidth access proposed by other countries, U.S. efforts with DSL and cable seem inadequate. In Canada, projects are beginning to deliver affordable access over fiber for institutions, small companies and even homes. A recent report issued by the Information Technology Commission of Sweden proposes that within the next five years broadband access at a minimum of 5Mbps will be possible for all facilities in Sweden using open dark fiber and the conduit owned by local governments. The Swiss National Science Foundation has plans to leverage developments in mobile communications and mobile devices to create a new paradigm for information society based on a new type of communications network without any dedicated centralized infrastructure.

While other countries are moving ahead with true broadband access solutions, the U.S. seems content with the relatively slow deployment of lower bandwidth DSL and cable modems for home and SOHO markets. Will Europe and Canada take the lead in the next 10 years in widespread development of optical and wireless broadband information and communications networks? I do not doubt that U.S. policy of unregulation (and a return to non-regulation) is effective in one aspect. In other aspects, however, this policy may be too gradual. I wonder if the U.S. requires a more drastic policy to maintain their lead in the information revolution in the future.

Japan's Telecom Future
I would like to turn now to conditions in Japan. Continuing deployment of ISDN, which is nothing more than a relatively low-speed service, and our lack of flat rate tolls are a major handicap for the spread of the Internet. In addition, the rate of diffusion of cable television in Japan is far below that of the U.S. and Canada, and many of NTT's local telephone circuits are poor quality. While the average distance from the central office to the subscriber is far shorter than that in the U.S., local copper line diameters are often smaller than those used in the U.S. Also, there are still many lines with paper insulation, which are not very suitable for DSL.

Ordinary DSL standards are not compatible with Japanese "ping-pong" ISDN. Therefore, Japan is faced with the real problem of having to adopt a different standard from the U.S. and Europe. This disadvantageous condition is reflected in the higher cost of DSL modems. In that sense, even if Japan rushes headlong into the full-scale diffusion of the Internet in the future, cable modems and DSL--favored in the U.S.--cannot be the only solution for Japan. As for ISDN, its limitations are obvious because it is capable of a maximum of 128 Kbps.

The road that Japan should follow is clear. Japan must exert every effort to create a new optical/wireless broadband network, with architecture different from the existing telephone network. It would be a condominium type network originating from each region, or in the form of what GLOCOM calls "Community Area Networks" (CANs). For this, the concerted efforts of local government, local communities, citizens, and corporations would be a more important condition than market competition. Also, to promote competition we should probably loosen regulations and adopt policy measures to further these efforts. It will be important to ensure "access to public roads for the installation of facilities" and access to other rights of way as stated in the Senate resolution.

My opinion is that rights of way in a broad sense, including spectrum, must be offered widely to those who wish to use them under standard rules and charges as the common property of the local community. The revenue from the rights of way would likely become an important source of income for local governments in Japan. I would like to see this become the next constitutional amendment.

To revisit an earlier point, the issue of reducing interconnection charges is an unproductive argument because the possibility of building new fixed-type local telephone facilities in the future, including switching systems, is almost nil. Thus, "the long-run incremental costs" for these facilities becomes an empty concept. However, using connectivity costs associated with IP telephony as the basis for calculating existing telephone connection charges will be equally unrealistic even if they are used in the future as the basis for "market based rates". With the Internet established as the new paradigm, it would be entirely expected that there would be no interconnection charges. Further, if NTT does not make rapid progress in actual cost reduction as profits decline from lower interconnection charges, they will be forced to consider other avenues for increasing revenue to avoid greater losses. One way currently being considered is to increase basic charges. In addition, NTT may also ask other carriers to contribute to the "universal service fund".

Time to End the Endless Debates
For more than a decade Japanese telecom policy has too often followed the lead of the U.S. Since the privatization of the Japan Telegraph and Telephone Public Corp. in 1985, which followed in the wake of the AT&T break-up in 1984, debate about how to do something similar with NTT continued intensely for more than 10 years. And while Japan debated how to break-up NTT, the opposite was occurring in the U.S. through increased mergers and acquisitions. Also, questions of unbundling of network elements and reducing interconnection charges had already been resolved in the U.S. Now this trend is being reversed, at least partly, toward unregulation.

Under pressure from the U.S., Japan is engaged in endless debate--based on the premise of the old model of telephone service and providing access to network elements--about how to reduce interconnection charges. Perhaps it will not be long before Japan too will have debates about unregulation. It would be sad if this is the prospect that awaits us.

What Japan needs now is not to follow in the wake of American policy. Rather, I think Japan must pay attention to the results achieved in the U.S. and consider either how to achieve the same results without falling behind or how to achieve superior results. I would like to see the U.S. support such efforts. As the U.S. Senate resolution states, "A sustained recovery of the Japanese economy is vital to a sustained recovery of Asian economies." For this reason alone, the Japanese economy must function as one of the primary engines of economic growth for Asia and the world.