In the first three years of the Bush administration, the United States dropped from 4th to 13th place in global rankings of broadband Internet usage. Today, most U.S. homes can access only "basic" broadband, among the slowest, most expensive, and least reliable in the developed world, and the United States has fallen even further behind in mobile-phone-based Internet access. The lag is arguably the result of the Bush administration's failure to make a priority of developing these networks. In fact, the United States is the only industrialized state without an explicit national policy for promoting broadband.
It did not have to be this way. Until recently, the United States led the world in Internet development. In the late 1960s and 1970s, the Department of Defense's Advanced Research Projects Agency conceived of and then funded the Internet. In the 1980s, the National Science Foundation partially underwrote the university and college networks -- and the high-speed lines supporting them -- that extended the Internet across the nation. After the World Wide Web and mouse-driven browsers were developed in the early 1990s, the Internet was ready to take off. President Bill Clinton and Vice President Al Gore showed the way by promoting the Internet's commercialization, the National Infrastructure Initiative, the Telecommunications Act of 1996, and remarkable e-commerce, e-government, and e-education programs. The private sector did the work, but the government offered a clear vision and strong leadership that created a competitive playing field for early broadband providers. Even though these policies had their share of detractors -- who claimed that excessive hype was used to sell wasteful projects and even blamed the Clinton administration for the dot-com bust -- they kept the United States in the forefront of Internet innovation and deployment through the 1990s.
Things changed when the Bush administration took over in 2001 and set new priorities for the country: tax cuts, missile defense, and, months later, the war on terrorism. In the administration's first three years, President George W. Bush mentioned broadband just twice and only in passing. The Federal Communications Commission (FCC) showed little interest in opening home telephone lines to outside competitors to drive down broadband prices and increase demand.
When the United States dropped the Internet leadership baton, Japan picked it up. In 2001, Japan was well behind the United States in the broadband race. But thanks to top-level political leadership and ambitious goals, it soon began to move ahead. By May 2003, a higher percentage of homes in Japan than in the United States had broadband, and Japan had moved well beyond the basic connections still in use in the United States. Today, nearly all Japanese have access to "high-speed" broadband, with an average connection speed 16 times faster than in the United States -- for only about $22 a month. Even faster "ultra-high-speed" broadband, which runs through fiber-optic cable, is scheduled to be available throughout the country for $30 to $40 a month by the end of 2005. And that is to say nothing of Internet access through mobile phones, an area in which Japan is even further ahead of the United States.
It is now clear that Japan and its neighbors will lead the charge in high-speed broadband over the next several years. South Korea already has the world's greatest percentage of broadband users, and last year the absolute number of broadband users in urban China surpassed that in the United States. These countries' progress will have serious economic implications. By dislodging the United States from the lead it commanded not so long ago, Japan and its neighbors have positioned themselves to be the first states to reap the benefits of the broadband era: economic growth, increased productivity, technological innovation, and an improved quality of life.
JAPAN'S HIGH-WIRE ACT
In the late 1990s, after a decade in the economic doldrums, Japan lagged well behind the United States in Internet access and broadband usage. But in mid-2000, Prime Minister Yoshiro Mori appointed the Information Technology Strategy Council, headed by Sony Chairman Nobuyuki Idei, which put together a bold plan to make Japan the "world's leading IT [Information Technology] nation" by 2005. Just as President Bush was taking office, a new Japanese "IT strategic headquarters," headed by the prime minister and including the entire cabinet, launched an "e-Japan strategy."
A central goal of that strategy was to bring better-than-basic broadband to 40 million of Japan's 46 million households within five years. The government hoped to make high-speed broadband available to 30 million households (through cable or digital subscriber lines [DSL], which use phone wires) and ultra-high-speed broadband connections to another 10 million (through fiber-optic cable). But even Japanese officials were skeptical about reaching such ambitious goals. And they understood that if they wanted even to come close, they would have to enlist the private sector and create the proper conditions.
The government quickly removed many regulatory obstacles. But because cable providers were mostly mom-and-pop operations in rural areas, officials realized that they would also have to create a highly competitive private-sector environment. So the telecommunications ministry came up with one of the most competitive regimes in the world: it compelled regional telephone companies to grant outside competitors access to all their residential telephone lines in exchange for a modest fee (about $2 per line a month). The antitrust authorities also ensured that these companies did not create obstacles for their competitors, helping provide a level playing field.
The results were extraordinary. Yahoo! BB, created by Masayoshi Son's venture-capital firm Softbank, and several other companies soon entered the DSL market. Yahoo! BB began offering high-speed service five times faster than current U.S. broadband for $22 a month. After aggressive marketing forced its competitors to meet Yahoo! BB's price, high-speed DSL subscriptions skyrocketed. By the end of 2002, such access was available to many more than the 30 million Japanese households the government had targeted. Within another five months, a greater percentage of homes in Japan than in the United States had access to broadband.
Thanks to the government's competitive framework, the speed of the DSL service offered also rose dramatically, from 8 megabits per second in 2001 to 12, 26, and 40 megabits today. (The typical U.S. broadband connection, whether DSL or cable, is still only 1.5 megabits per second or slower.) Meanwhile, the price of monthly subscriptions remained stable, even for 26-megabit access speeds, at about $22 per month -- by far the lowest price in the world. By September 2004, 15.3 million Japanese subscribed to high-speed broadband. Moreover, for an additional $5 per month, users of Yahoo! BB can also have Internet telephone service. One in every 25 telephone calls in Japan is now made over the Internet, and the number keeps growing.
Meeting the e-Japan strategy's second goal -- making ultra-high-speed access (up to 100 megabits per second) available to ten million Japanese households -- proved more difficult. Such connections permit real-time video telephoning and video conferencing, telecommuting, and rich multimedia options such as digital high-definition television, interactive games, and five-minute movie downloads (instead of the short, jerky video streaming that Americans are used to). But data cannot be transmitted at such speeds through existing phone lines, and new fiber-optic cable had to be laid throughout Japan. Having decided that those lines, too, should be open to competition, the Japanese authorities set out to devise significant incentives to persuade Japanese companies to invest in new ultra-high-speed cable, especially in rural areas.
The government used tax breaks, debt guaranties, and partial subsidies. It allowed companies willing to lay fiber to depreciate about one-third of the cost on first-year taxes, and it guaranteed their debt liabilities. These measures were sufficient to ensure that new fiber was laid in cities and large towns, but in rural areas, municipal subsidies were also needed. Towns and villages willing to set up their own ultra-high-speed fiber networks received a government subsidy covering approximately one-third of their costs, so long as those networks, too, were open to outside access.
These incentives created the right environment for the rapid deployment of fiber networks. Again, other companies decided to compete with regional telephone companies. The first, Usen, a nationwide distributor of background music with its own fiber network, was later joined by electric power companies. The resulting competition quickly drove the price of an ultrafast fiber connection down to $30 to $45 per month.
By the end of 2002, ultrafast fiber connections were available to more than ten million households in Tokyo and Osaka; a primary goal of the e-Japan strategy had been met. But the program -- and the government's tax incentives - had also called for fiber lines to run directly to homes and offices, and those connections proved economic only in densely populated cities. In less settled areas, the government agreed to provide tax incentives for fiber taken only as far as neighborhoods, leaving it to individual users to decide how to connect. Some have chosen -- and paid for -- a direct fiber connection; others have opted for a cheaper but slower wireless connection. By mid-2004, ultra-high-speed broadband was available to more than 80 percent of Japan's citizens. With more than two million subscribers, it can be said to have gone mainstream.
Fiber deployment is still moving quickly, and by the end of the year, ultra-high-speed access will be available to virtually all Japanese either directly or in their neighborhood. The program has been so successful that the Japanese government has already set its sights higher: in mid-2003, it decided to move beyond promoting access to ultra-high-speed broadband to encouraging its use.
ON THE FRITZ
So far, no one in the Bush administration has offered a vision nearly as compelling as Japan's. Although Michael Powell, the former chairman of the FCC, spoke eloquently about the benefits of the coming "digital broadband migration," he suggested no date for arrival in the promised land. Moreover, he measured U.S. broadband progress by the exceptionally slow 200-kilobit-per-second standard -- about one-hundredth of the speed of typical broadband in Japan today. According to that minimal standard, the United States has made some progress: by mid-2004, more than 30 million American homes and offices had signed up for basic broadband. But the service is expensive, very slow, and rather unreliable. And despite these limitations, the Bush administration has made little effort to encourage cheaper and more robust high-speed broadband or to promote what many agree should be the model for the future: a vast network of ultrafast fiber connecting homes, offices, and neighborhoods.
Without vision or leadership, U.S. broadband policy drifted during the Bush administration's first two years. The FCC tended to other matters. The Department of Commerce insisted that the market, not the government, should drive the rollout of broadband. Meanwhile, regional telephone companies relentlessly tried to reverse some of the promising measures that had been taken under President Clinton. Continuing efforts they had launched after the 1996 Telecommunications Act was passed, they lobbied legislators and sought court decisions to overturn regulations that had forced them to open their residential telephone lines to competitors.
Powell seemed not to mind this challenge; he preferred a somewhat different approach anyway. He backed promising new technologies and appeared less interested in the idea of promoting DSL competition for residential telephone lines, even though the strategy had quickly boosted access speeds and lowered prices in Japan and elsewhere. Instead, he favored pitting the cable television industry against the regional telephone industry.
Although in theory the strategy was viable -- telephone and cable lines run in front of more than 75 percent of U.S. homes, and with some technical upgrading, both can provide basic or high-speed broadband -- many opposed it. Among the critics of the multiplatform approach were Powell's predecessors at the FCC, who had done their utmost to open residential telephone lines; many economists, who were distrustful of duopoly competition; and consumer groups. Firms that were already competing or that wanted to compete with regional telephone companies in providing DSL service disagreed, too, as did those that coveted access to cable television lines. Some even claimed that this approach violated the 1996 Telecommunications Act, which, they argued, required the sharing of residential telephone wires.
Still, when the FCC got around to reviewing broadband policy in February 2003, it made convoluted decisions that left only the multiplatform approach. Firms that were competing with regional telephone companies to offer high-speed DSL service over telephone lines would have only three more years of access. More significant for the long run, the regional telephone companies would not have to share with outside competitors the ultra-high-speed fiber lines they laid. The following year, moreover, at the urging of regional telephone companies, a court reaffirmed an earlier ruling that these companies need not share their residential lines with DSL competitors. Although many expected an appeal, higher levels of the administration chose not to challenge the decision. Thus, broadband competition over residential telephone lines was effectively killed. A proven strategy had been lost.
Unfortunately, vigorous multiplatform competition is unlikely to emerge soon. True, there are signs of competition between the cable-modem broadband offered by cable television companies and the DSL service offered by telephone companies. Comcast plans to provide reliable Internet-based telephone service by doubling the speed of its broadband offerings from 1.5 megabits to 3 megabits per second over the next three years. Verizon and SBC Communications have dropped the cost of their broadband service to about $30 a month. And to compete directly with cable, some phone companies have begun to talk of developing their own Internet telephone service and providing higher broadband speeds to deliver video.
But these new services will probably appear only slowly, and competition between the telephone and cable companies will remain limited. The reasons are simple: cheap, high-speed broadband would lead to widespread use of Internet telephones and thus threaten the phone companies' lucrative voice-telephone business, and more inexpensive broadband would multiply outside video and movie offerings and endanger the cable companies' profitability. So, although both the telephone and cable companies could provide cheap, high-speed broadband if they chose to, they are not rushing to develop it.
The lack of strong incentives to encourage competition has, in other words, doomed broadband in the United States to remain much slower and more expensive than in Japan. Over the next five years, service is likely to get only marginally faster and cheaper. Meanwhile, at current transmission speeds, the next "killer" application -- Internet telephone service -- will remain shaky and unreliable.
The development of ultra-high-speed fiber broadband service, which is just beginning to appear in the United States, will also lag. Barely more than 600,000 U.S. offices and homes had fiber connections at the end of 2003. Verizon plans to bring fiber to 3 million of the United States' 115 million households by the end of this year, with speeds ranging from 5 to 30 megabits per second. SBC Communications, which dominates the Midwest and Southwest markets, and BellSouth, the leader in the Southeast, are also laying fiber, although at a much slower rate. But they plan to stop the work after spending about $10 billion (the estimated cost of bringing fiber close to about 10 million U.S. homes and offices) and then examine whether further investment is justified. As a result, the pace of rollout will be slow. And the emergence of the substantial market needed to inspire innovative new products and services for those with fiber Internet access remains years away.
PLAYING PHONE LAG
The United States is even further behind Japan in wireless, mobile-phone-based Internet access, even though that platform is increasingly versatile and valuable. More and more, mobile phones can be used for tasks traditionally performed on computers. Except for the most office-oriented applications, such as word processing, spreadsheets, and presentation software, mobile phones will soon be used for nearly everything. In fact, many, including the Japanese, are already planning for a convergence of wireline and wireless technologies. By 2010, it is expected that such "ubiquitous networks" will permit Japanese to access the Internet at high speeds from a desktop, a laptop, a hand-held personal digital assistant, or a mobile phone.
Japan now has a commanding lead in mobile-phone Internet technologies and usage. With a nationwide cell-phone infrastructure in place by the mid-1990s, Japan began the shift away from voice services to Internet data services in early 1999. Then NTT DoCoMo introduced the "i-mode" service, providing e-mail and customer access to over 60 Web sites especially created for mobile-phone use. These sites offer news, financial services, weather, personal ads, games, and much more. (This service was recently introduced as "m-mode" in the United States.) Competitors soon emerged, and customer response was stunning. By December 2004, total mobile-phone subscriptions had reached 83.5 million in Japan (representing more than 60 percent of the population), of which more than 72 million included Internet services. The lesson the NTT DoCoMo leadership took from this experience was that if you develop a new technology and market it, consumers will buy it.
Following this philosophy, in October 2001, NTT DoCoMo launched a third-generation videophone service. By December 2004, thanks to thriving competition, Japanese videophone subscriptions had reached nearly 26 million and were growing by nearly 190 percent a year. As expected, this new market prompted notable mobile-phone innovation such as global-positioning-linked advertising, television reception, and music videos. Now Japan is testing fourth-generation, high-speed broadband phones that can support high-definition-television reception, movie downloads, more sophisticated games, and other multimedia applications.
The Japanese government played a critical part in these developments. It made well-considered and timely decisions to allot cost-free spectrum for each new mobile-phone generation. In so doing, it gave up badly needed revenue, but it retained full control over the terms of licensing and the flexibility to reassign spectrum according to future technological developments. In 2007, the government is expected to announce new spectrum allocations for the fourth-generation broadband mobile phones planned for 2010. Meanwhile, to protect consumers, the government has set important conditions before granting a service license, insisting that a carrier's network cover a certain area of the country and guarantee a certain level of service (with minimal dropped calls or interference, for example).
By contrast, U.S. mobile-phone policy was born of a colossal blunder from which the industry has yet to recover fully. In the early 1980s, after the management consultancy McKinsey estimated that there would be little demand for mobile phones and a small prospect of profitability, the FCC carved the United States into 734 tiny mobile-phone districts. It handed out two provider licenses in each district: one automatically went to the regional telephone company, and the other was drawn by lottery. The resulting infrastructure was cripplingly fragmented. It could not support nationwide calls, and inefficiencies and expensive connection rates translated into sky-high charges for customers.
Twenty years later, the Clinton administration made a belated effort to encourage nationwide cellular networks. The government opened up enough spectrum for six nationwide networks and invited bids. Thanks to an imaginative on-line auction, it had sold off the spectrum for $7.7 billion by early 1995. Although the networks that entered the market still struggle to offer consistent quality, competition among them sharply reduced the price of mobile-phone service and spawned millions of new customers.
Since the Bush administration took office, however, the FCC has only tinkered with spectrum policy around the edges. It has allowed companies to trade bits of spectrum to round out their infrastructure and opened modest amounts of spectrum to new wireless technologies such as WiFi and WiMax. Meanwhile, although the number of would-be national carriers dwindled from six to four and they expanded their infrastructure, U.S. mobile-phone service remains awful by European, let alone Japanese, standards. U.S. mobile phones can take digital pictures and connect to the Internet, but the cellular infrastructure is so spotty that even in large cities calls from an ordinary wireless phone may not go through. Sadly, U.S. mobile-phone competition is still based on price and the extent of a company's coverage rather than the kind of advanced data services available in Japan and elsewhere.
In 2004, third-generation mobile service came on the market in selected U.S. cities. As in Japan, two somewhat different technologies are being used, both of which require upgrading the existing infrastructure. For the time being, third-generation mobile-phone service is available in only eight cities. (The much slower, older service can be had in several others.) Although the FCC has provided some additional badly needed spectrum, the third-generation cellular infrastructure remains painfully inadequate: most of the country has no service at all. Meanwhile, the FCC has announced that it will auction third-generation spectrum "as early as June 2006." Plans for fourth-generation mobile service in the United States are well beyond the horizon.
GETTING BACK ON-LINE
The United States is losing considerable ground to Japan and its neighbors, and they will be the first to reap the economic benefits of these technologies. It is these countries, rather than the United States, that will benefit from the enhanced productivity, economic growth, and new jobs that high-speed broadband will bring. In 2001, Robert Crandall, an economist at the Brookings Institution, and Charles Jackson, a telecommunications consultant, estimated that "widespread" adoption of basic broadband in the United States could add $500 billion to the U.S. economy and produce 1.2 million new jobs. But Washington never promoted such a policy. Last year, another Brookings economist, Charles Ferguson, argued that perhaps as much as $1 trillion might be lost over the next decade due to present constraints on broadband development. These losses, moreover, are only the economic costs of the United States' indirection. They do not take into account the work that could have been done through telecommuting, the medical care or interactive long-distance education that might have been provided in remote areas, and unexploited entertainment possibilities.
The large broadband-user markets of Northeast Asia will attract the innovation the United States once enjoyed. Asians will have the first crack at developing the new commercial applications, products, services, and content of the high-speed-broadband era. Although many large U.S. firms, such as Cisco, IBM, and Microsoft, are closely following developments overseas and are unlikely to be left behind, the United States' medium-sized and smaller firms, which tend to foster the most innovation, may well be.
The Japanese and the South Koreans will also be the first to enjoy the quality-of-life benefits that the high-speed-broadband era will bring. These will include not only Internet telephones and videophones, but also easy teleconferencing, practical telecommuting, remote diagnosis and medical services, interactive distance education, rich multimedia entertainment, digitally controlled home appliances, and much more.
Given these costs and losses, it is clear that broadband is critically important to the U.S. economy and the United States' international competitiveness and that it must become a national priority. In the run-up to the election in November, President Bush finally addressed the issue, promising the electorate "universal, affordable access" to broadband technology by 2007 and "plenty" of carriers to choose from "as soon as possible thereafter." To reach these goals, he expressed confidence in new broadband service over power lines, promising wireless technologies, such as WiFi hotspots and longer-distance WiMax, and unspecified tax credits.
But real progress will require more than these measures. To move forward, the administration should quickly take two steps. First, it should explain clearly the profound ways in which broadband will change work, learning, and leisure in the United States. Identifying such substantial benefits would energize providers and encourage potential users to get the most from the Internet. It would also give the private sector confidence in the nation's direction and a degree of business certainty.
Second, the administration should push the President's Information Technology Advisory Committee (PITAC), a group of private-sector IT leaders and academics, to play a key leadership role in advancing broadband deployment. Involving the private sector and prominent academics in broadband leadership is essential given the pace of technological advance and today's dynamic business environment.
One of the PITAC's first tasks should be to set out bold long-term goals for the deployment of broadband in the United States, carefully distinguishing three different levels of service: basic broadband (at 1.5 to 3 megabits per second), for slow downloads from and uploads to the Internet and Internet telephones; high-speed broadband (at 10 to 30 megabits per second), for Internet reception of digital high-definition television and other video uses; and ultra-high-speed fiber broadband (at 100 megabits per second), for the highest-end applications.
The PITAC should consider how to redeem President Bush's pledge to provide, by 2007 (or 2010, at the latest), basic broadband access to all Americans at an affordable price ($20 to $25 per month should be the goal). To reach everyone, the effort would require developing a combination of technologies: wireline, wireless, and satellite. The United States' vastness no doubt complicates the task, but it is no excuse for not undertaking the job. (Canada, the world's second-largest state, also ranks second in global broadband connectivity.) If necessary, tax credits should be granted to companies that help reach rural and underserved areas.
By 2010, the PITAC should also aim to make available high-speed broadband access to two-thirds of all U.S. households for $30 to $35 per month. The key to reaching this goal is the government's taking the lead in creating a strongly competitive environment for DSL, cable, power line, and newer wireless broadband technologies. The more these technologies compete among themselves, the sooner Americans will have access to faster, cheaper broadband service. And with enough competition, there should be no need for government financial incentives.
The PITAC should also do its best to promote ultra-high-speed fiber access for one-third of all U.S. households at $40 to $45 per month by 2010. It should use its convening power to bring to the table all the stakeholders in the millions of miles of unused fiber that run below U.S. city streets. The purpose of such discussions would be to encourage the widespread use of existing fiber by analyzing the reasons for its current disuse and seeking ways to make it viable. The PITAC might also recommend legislation to permit the National Science Foundation to provide matching grants to bring fiber to the campuses of colleges and universities across the country. This program could be modeled on the highly successful National Science Foundation Network (NSFnet) project that brought the Internet to campuses in the 1980s.