Keeping the U.S. Computer and Communications Industry Competitive: Convergence of Computing, Communications, and Entertainment (1995)

Digital convergence introduces new problems into the ongoing debates over international competitiveness, telecommunications policy reform, and the evolution of the information infrastructure. Policymakers are challenged to balance the different priorities, national interests, and legal and regulatory traditions of the telecommunications, computer, and entertainment industries. At the same time, there is pressure on government to act—both to relax existing laws and rules and to enact new ones—because information technologies are evolving more rapidly than ever before.

Central to the policy debate are the prospects for an enhanced information infrastructure, which galvanized private- and public-sector attention during 1993-1994.¹ Legislative efforts during that period constituted the first major attempt since passage of the Communications Act of 1934 to alter the goals and frameworks for state and federal regulation and the constraints on business entry by common carriers provided by the Communications Act of 1934 and the Modified Final Judgment codifying the breakup of AT&T.² Although telecommunications reform legislation failed to pass in 1994, new legislation was introduced in 1995. Contention about the specifics suggests that whatever legislation does pass will launch a process in which various parties will seek to fix and fine-tune those aspects that affect them.³

Developments over the ensuing months indicate that comments at the colloquium were symptomatic of the kinds of concerns being expressed in multiple forums. They reflect the political process that has driven legislative efforts to overhaul the legal and regulatory framework for telecommunications. For example, Samuel Ginn, speaking then as the head of a regulated common carrier

(Pacific Telesis), was perhaps the most forceful in criticizing the constraints on telecommunications firms, contending that "public policy and regulatory policies will be the most significant factor in determining whether these kinds of services can be made available to the public, not technology." Ginn asserted that public policy and regulation will be the most important factor—and one of the toughest challenges—in making advanced information services available to all. Similarly, Ameritech's Richard Notebaert said government policy is the "major speed bump" inhibiting the application of technology to solve societal ills. Beneath these general statements lie concerns about the mechanisms (e.g., delays and uncertainties associated with multiyear proceedings) as well as the objects (e.g., ability to enter certain markets, terms of participation in certain markets, and pricing constraints) of regulation. Those concerns were aired universally in a series of late-1994 interviews (Appendix C lists those interviewed), suggesting that between 1993 and 1995 both interest in market entry and concern about constraints had grown among the various players in the digital convergence arena.

Despite concerns about the impact of regulation on the cost and profitability of a given investment, comments by several participants suggest a broad recognition that absent government intervention, the owners of key capabilities control timing and access. Thus Ginn observed, "It's not clear to me, given the way the markets are developing, that there is an opportunity here to serve all elements of society unless you have intervention in terms of social policy."

The proliferation of public interest advocacy groups in the 1993-1994 period attests to a broadening concern about how the public policy agenda for information infrastructure is being set and how various objectives are weighted when decisions are made. Yet another face on this concern was provided by the change in the Congress with the 1994 elections. This process of politicization also reflects a recognition that the distribution of benefits from digital convergence will vary over time: costs come down and access can broaden over the long term. The question raised by public interest advocates is, How long will and should be the wait for mass access to technology's benefits? The question emphasized by business is, Who pays to direct or accelerate receipt of benefits that do not derive directly and naturally from the evolution of various markets?

Because of leadership in the fundamental markets associated with it, digital convergence, in the judgment of the colloquium steering committee and other observers, should contribute significantly to the economy and international competitiveness of the United States. U.S. companies lead the world in most aspects of computing and communications technology and entertainment programming; in addition, innovation-stimulating market competition is more prevalent in the United States than in rival nations, and English increasingly is becoming the international language of commerce and entertainment (Memmott and Maney,

1993). David Nagel asserted that, despite recent erosion of U.S. dominance in computers, "it's clear still that nearly all of the interesting innovations, both in computing and in telecommunications, are being driven in the U.S. industrial base." Whether this dominance is sustainable, Nagel said, depends on whether U.S. firms have learned the lessons of the past and can apply and profit from their own innovations.

Entertainment adds a new and unknown variable to the mix: it is seen as a source of economic benefit to its purveyors, but unlike computing and communications tends not to be seen as an "intermediate good" or vehicle to broader benefit except as an item of mass consumption. There are therefore more assumptions than facts when it comes to assessing entertainment's contributions to competitiveness. In a late-1994 interview, Peter Cowhey, formerly a professor of international relations and political science at the University of California at San Diego, noted that entertainment products are receiving more attention in trade policy, with leaders in some countries showing growing concern about the prevalence and impact of U.S.-originated programming. It may require explicit action, he suggests, to point out and ensure that the expanding, advancing global information infrastructure will provide capacity for distributing cultural outputs from everyone—it is not an exclusive vehicle for distributing U.S. content, although global distribution outlets (of which Music Television may be a crude forerunner) are likely to emerge, and it broadens the market for cultural outputs generally (i.e., it makes the pie larger).

A sister issue, notes Cowhey, is the issue of ownership of media. Current U.S. policy restricts foreign ownership of broadcast media to 25 percent,⁴ and some countries are introducing reciprocal restrictions or program quotas to limit U.S. entry, in particular.

The National Telecommunications and Information Administration (NTIA) also warned in a 1993 report that the United States "cannot afford to be complacent about the success of U.S. media firms in international markets. Recent regulatory and technological changes require U.S. policy makers to continue to adapt in order to promote the development of international mass media markets that are open and competitive—the type of markets in which U.S. firms historically prosper" (NTIA, 1993). The NTIA made a series of recommendations designed to foster economic growth in the mass media industry, not only in the United States but also—because of the global nature of the industry—worldwide.

The United States has provided fertile ground for development and application of a number of services that provide a working foundation—in the form of corporate learning and consumer familiarity and expectations—for digital convergence. As Michael Borrus has argued, technical progress depends on an economy's potential for "learning by doing and by using" (Borrus, 1993). See Box 4.1. Use of digital technologies has not only benefited organizations and individuals within the United States, but has also provided the basis for the development of global enterprise, including global service industries that tap and

deliver knowledge and talent wherever it resides or is needed (CSTB, 1992, 1994a). This phenomenon has itself resulted in a refocusing of trade policy and trade negotiations on trade in and conditions for service industries as opposed to physical product or commodity distribution. Cowhey remarked on these developments in a late-1994 interview:

Cowhey noted that the large and diverse installed computing base has led to an artificially large U.S. lead in digital convergence.⁵ He observed that it is reasonable to expect greater diffusion—and associated benefits—of personal computers and other systems in at least the industrial nations. Moreover, explained Fuller in another interview, "By its very nature, an information utility or an internetwork will be global as well as national in scale, and issues of geography will become less important." As Cowhey noted, the 1993-1994 public statements of President Clinton and Vice President Gore focused attention on information infrastructure across the industrialized world, making it folly to assume that only the United States would place emphasis in that area. Indeed, the "G7" industrial countries held their first ministerial summit devoted to a single industry when the European Union hosted the February 1995 "G7 Ministerial Conference on the Information Society." The European Union particularly championed the event in order to build political support for European regulatory reforms necessary to catch up with the United States. Box 4.2 lists 11 multinational pilot projects agreed upon at the G7 summit, which also resulted in agreement on several high-level principles:⁶

• Promoting fair competition;

• Encouraging private investment;

• Defining an adaptable regulatory framework; and

• Providing open access to networks while

• Ensuring universal provision of and access to services;

• Promoting equality of opportunity to the citizen;

• Promoting diversity of content, including cultural and linguistic diversity; and

• Recognizing the necessity of worldwide cooperation with particular attention to less developed countries.

European nations have recently moved to launch their own video-on-demand trials, often drawing on U.S. technology. The cable television business is more advanced in the United States than elsewhere: U.S. firms are supplying cable equipment and programming worldwide. Further, as George Gilder observed, "No other country in the world has broadband connections to 60 percent of homes or passing 90 percent of its homes." Although there have been publicized instances of high-level capabilities in specific locations overseas, the actual average capability available to almost all households appears higher in the United States than elsewhere. On the other hand, foreign suppliers of consumer and other electronics, such as Thomson Consumer Electronics, are a significant presence in the U.S. market. Although slower European progress has been attributed to the high cost of installing high-bandwidth communications facilities, advances in data compression and other technologies for more fully exploiting copper wire and advances in computer systems have been cited as encouraging European

efforts. Moreover, cellular telephony has spread relatively rapidly in Europe, aided in part by agreement on a supporting standard (GSM).

Tempering the discussion of competitive advantage and disadvantage, Borrus suggested that evolving international business relationships may become more complex and cooperative. He noted that the United States controls the ''soft" side of the electronics industry—from design, architecture, and marketing to software and systems integration—but that key hardware technology and manufacturing activities have been migrating to Asia. "No one geographic region, and certainly no individual company, has all of the necessary know-how to develop this next generation of electronic systems," Borrus said. Yet the persistence of U.S. leadership in software and U.S. reliance on offshore manufacturing raise questions

about what trends and indicators are most reliable. Moreover, it may be that digital convergence represents an opportunity to reclaim areas lost to foreign competitors, such as consumer electronics, to the extent that such areas become more computer-like.

Eli Noam has said that international asymmetries hamper the global system of systems, which will function effectively only if it is competitive at each stage, and if regulation establishes nondiscrimination. At present, the information infrastructure stands at uneven levels of development in technology and policy across the globe; most countries still have an established monopoly communications provider, and outside of the United States monopoly remains the dominant communications paradigm. Government policies and limited competition abroad

may be a greater impediment than available technology; certainly, those factors have limited both deployment and experimentation with computing and communications technologies in the past, to a far greater extent than have policies in the United States. Telecommunications reform is emerging in many countries (Hudson, 1994; Lavin, 1995), and considerable discussion has surfaced over the desirability of more market-oriented regimes (Bangemann et al., 1994).

In a late-1994 interview, Cowhey seemed to anticipate the 1995 "G7 Ministerial Conference on the Information Society" by describing the importance of policy relating to broad social objectives as a factor in the international competitiveness dynamic:

Borrus also pointed to the need to address international asymmetries in technology access. He has warned that the U.S. supply base is likely to be more accessible to foreign rivals than is the Japanese market, and that software skills and technology can be appropriated more easily than can manufacturing components and expertise (Borrus, 1993). Borrus outlined the problem of how to assure equitable access to both U.S. and foreign companies. It is easy to invest in a U.S. company and thereby gain access to know-how,⁷ but very difficult to gain similar access to a Japanese company, for example. The Japanese have discriminatory public procurement policies (COC, 1993), and their keiretsu system of business alliances establishes preferential sales, supply, and capital allocation arrangements that impede market access by new entrants.⁸ Also, there are far fewer publicly held companies in Japan (2,000 as of 1987) than in the United States (24,000; Gerlach, 1992). Borrus anticipated a shift in U.S. trade policy to deal with the access problem. Experiences in the semiconductor arena illustrate the possibilities; see Box 4.3. In pursuing open markets worldwide as a trade policy objective, suggests Cowhey, the United States must attend to concerns in other countries that they not open their markets at a point in time when the United States appears to be closing its own.

Another, more perplexing problem, according to Borrus, is to identify the foreign technologies and know-how that should reside in the United States, and to find ways of moving them here. An example is flat panel displays, used in applications such as laptop computers. Flat panel displays were invented in the United States over 35 years ago, but Japanese companies now control over 95 percent of the burgeoning $3.5 billion market and outspend U.S. firms on display

research by more than 20 to 1. Various means have been explored for strengthening the U.S. base in display technologies. "Dumping" duties were imposed on imported Japanese displays in 1991, but that trade policy has had mixed results. The Advanced Research Projects Agency launched a research initiative to promote innovation. A broader approach may be called for. However, the Republican congressional leadership has expressed strong doubts about such programs, worrying that they constitute little more than subsidies for large electronics firms.

Symmetrical international access to technology—including increased U.S. access to certain technologies—will not emerge unaided. There is no unanimity on the precise mix of policy. For example, Gilder has criticized the semiconductor agreement praised by Borrus. But Borrus captured the terms of the policy discussion when he wrote that trading blocs need "(1) to agree on a set of principles that endorse reciprocal access to regional markets, investment opportunities, and supply base technologies; (2) to negotiate for tangible results that mitigate the disruptive impacts of domestic practices that violate agreed norms of behavior; and (3) to develop new multilateral institutions for coordinating bilateral regional moves" (Borrus, 1993).

Cowhey has noted that trade pacts have limited effect, in part because so much of world commerce is now determined by investment flows and global production strategies of multinational firms. This has heightened the importance

of emerging industrial alliances in international competition. For example, IBM Corporation has established a joint factory with Japan's Toshiba using the latter's liquid crystal display technology, and the two companies are working with Siemens of Germany to develop an advanced (256-megabit) dynamic randomaccess memory (DRAM) chip (Cowhey and Aronson, 1993). Of particular significance is the location of development teams at IBM's facilities in New York State. Such alliances permit the development of common, global infrastructures for technology development and allow firms to reduce the costs and risks of producing extensive product lines, Cowhey has noted (Cowhey and Aronson, 1993). National technology and trade policies, especially those for the information infrastructure, have to be evaluated in light of their impact on these alliances because much of the world's information infrastructure will be developed by such alliances.

How to design policies to stimulate innovation and promote competitiveness remains a subject of contention. A Council on Competitiveness project recently concluded that, if policymakers determine that U.S. development of a certain technology is in the national interest, "then they must create an integrated, systematic approach to the industry which includes capital market reform, the establishment of consortia, [and] support for the underlying technology infrastructure, along with trade policy" (COC, 1993). That report argues that policies to stimulate technological innovation in the United States are a necessary complement to trade policy.

Unfortunately, achieving an integrated, systematic approach is elusive. Experience in economics and industry suggests that chaos and conflicting efforts do better. The high-definition television (HDTV) situation, subject to a variety of interpretations, is a case in point. See Box 4.4.

One area where there appears to be consensus on the need for government action is international aspects of intellectual property protection, a source of growing concern as the information infrastructure becomes more global. The federal government has recognized the need for stronger intellectual property laws worldwide and in recent years has promoted international standards in this area. For example, the United States joined the Berne Convention for the Protection of Literary and Artistic Works⁹ and has bargained with other countries through the General Agreement on Tariffs and Trade and other regional and bilateral negotiations. Since digital transmissions are increasingly global, and since our evolving national information infrastructure is effectively part of a global information infrastructure, some mechanisms will be needed to handle the problem of differences among countries as to what constitutes a copyright infringement. For example, the concept of a parody as a fair use does not exist in all countries.

The tension between cultivating an integrated approach and leveraging the unexpected fruits of competition and some degree of market chaos is epitomized by policy debates relating to the evolving information infrastructure.

An overview of the U.S. approach to managing digital convergence was provided by Cowhey. The forces driving the domestic market are the pursuit of profit, political opportunity, long-standing industrial and technological structures, and cultural values, he said.

But foreign rivals have their own difficulties, Cowhey said. Europeans have encouraged technical cooperation among producers to achieve early integration of knowledge and effort. They have discouraged new entrants in the market and have not encouraged price cutting. The growth of interactive information industries is constrained by the slow changes in pricing and service practices of the telephone companies, and hardware development is sluggish because networks are providing the wrong types of incentives. Japan, meanwhile, has introduced some competition into its telecommunications industry, but the new telecommunications carriers are owned by equipment suppliers and, to some extent, by their large corporate customers. Moreover, public policy still tilts toward subsidizing equipment suppliers at the expense of users. And regulators have not allowed the dominant telephone companies to get into cable television.¹⁰ But even this barrier may fall, as Bell Atlantic Corporation's progress in pursuing judicial and

regulatory authorization to offer commercial video programming over its telephone distribution system attests (Landler, 1995; Carnevale, 1994).

"We're living through a great moment of experimentation here, and part of our problem in making public policy is that it's not self-evident which bet is right," Cowhey said. Both the European Union and Japan decided in 1994 that their approaches had to change. The European Union required the introduction of competition in voice telephone services by 1998. Japan reformulated many regulations that had hindered competition and actively debated breaking up NTT, its dominant phone company.

The United States also wrestled with changes in 1995. It was trying once again to legislate permission to allow the regional Bell operating companies, cable systems, and long-distance carriers to enter each other's markets. It also promised (in a speech by Vice President Gore at the 1995 summit of the G7) to remove the 25 percent limit on foreign ownership of spectrum, a limit that the foreign firms cited as a bar on their market access to the United States. The Vice President promised to achieve this goal by either legislation or regulation by the end of 1995. The one condition was that restrictions on foreign investment would be removed only for countries that liberalize their own markets. In short, this was an offer to liberalize on a bilateral basis with like-minded countries.

The same speech by the Vice President affirmed an even more sweeping goal, that is, success in the General Agreement on Trade in Services (GATS) trade negotiations on liberalizing basic (i.e., voice) telecommunications services scheduled to conclude by April 1996. If successful the GATS pact would make true competition in telecommunications services into a binding trade obligation for major industrial and some developing countries. Such a pact was considered unthinkable even 10 years ago.

The promise of information infrastructure clearly depends on considerable investment by private parties. At issue are not only competing technological and business approaches, but also the degree to which facilities and services offered by different parties can interoperate. Technical features—including those relating to openness and symmetry in communications, equity in access to infrastructure, and who gets to enter what markets—are among the key issues that informed colloquium discussions and more recent debates. These features have figured in the development of regulatory reform legislation, and they are also at the heart of initiatives relating to standard setting for voice, video, data, and multimedia services as well as to the formatting and transfer of various kinds of information and documents. For example, the Cable Act of 1992 provided for measures to foster greater compatibility between consumer electronic devices and cable television systems, resulting in digital convergence-related standard setting by the Federal Communications Commission (FCC) and industry. Also, the administration sponsored a mid-1994 workshop to identify key issues and options for facilitating standards development. But it is an open question as to how far the government can or should go in promoting the kinds of standards

needed for digital convergence; its efforts are likely to focus on points at which different networks and other infrastructure come together.

Colloquium participants suggested that government policies with respect to digital convergence should be guided by two general principles: first, that relaxation of traditional forms of regulation and increased market competition are beneficial in general; and, second, that there will continue to be a need for government intervention, but in different forms and areas than in the past. For example, many referred to growing (perceived) needs to protect intellectual property rights and to ensure that those who control a conduit not have a monopoly on the content that goes through it. Behind such comments lies the recognition that a growing portion of the investment in and returns on information infrastructure relate to information or content. Although most of the policy discussion reflected conditions in the telecommunications industry in particular (and within that the telephony component), many comments were relevant to information industries in general. On the other hand, there is little foundation in terms of regulation and other legal constraints affecting entertainment technology, and possibly even less constraining the computing industry.

The first principle may be obvious, because the benefits of market competition are well known. Noam and Cowhey have documented, in separate papers (Noam, 1995), the benefits derived specifically from the breakup of AT&T. Noam, in an analysis of the accuracy of various predictions concerning deregulation, has concluded that the doomsayers generally were wrong—that few benefits attributed to the monopoly system were lost. He found that proponents of deregulation were right more often than were pro-regulators, although liberalization evoked new issues that have been addressed effectively by regulation. In addition, Noam has concluded that the telecommunications sector is more dynamic and innovative as a result of deregulation.

Cowhey also has pointed to the benefits of liberalization, citing a number of statistics: since the breakup of AT&T, long-distance service cost has declined about 50 percent, the average household has seen a reduction in service costs, and the productivity of telephone companies has climbed an estimated 40 percent, all while these companies have introduced greater innovation in services and increased flexibility for users (Cowhey, 1990).

The second principle—concerning the changes in policy needs—requires thorough exploration. The breakup of telecommunications monopolies and the emergence of new media are having two primary effects on the policy-making landscape, Noam explained. First, there is less need for regulation to balance the power of small users facing giant, often monopolistic suppliers. Second, price and quality regulation also becomes less important as a systems integration industry emerges and begins to compete for consumers.¹¹

According to Noam, government should not react to this transformation by picking favorites—that is, by influencing or impeding change to favor any particular technology or carrier. Expanded Robert Johnson in a late-1994 interview, even the best-intended government intervention can distort the market, interfering with market demand and entrepreneurial creativity. Noam and others generally agreed that policymakers may need to promote rules of the road for the "network of networks," such as in interconnection and common carriage. Policymakers need to assure access to scarce facilities that could become "bottlenecks" in the marketplace, and they need to address international access asymmetries, he said.

As noted earlier in this report, with the development of specialized networks and the evolving need for content-selection aids, the economics of basic carriage become less attractive. The real advantage to a network owner increasingly will accrue from diversity in programming and services, and so there will be expanded opportunities for niche marketing; a network owner will become, in effect, a manager of a shopping mall, housing many "boutiques."¹²

In their quest to enter new markets, telephone companies have contended that substantial traffic must be generated to justify the building of fiber networks, that voice alone will not generate such volume. The NTIA, in a 1993 report, maintained that freeing telephone companies to provide cable programming would stimulate competition in the video market and provide incentives for infrastructure development, and, furthermore, any resulting increased demand for video programming could stimulate international trade (NTIA, 1993). Wildman has argued that, given the substantial costs of desired network upgrades and the nation's dependence on private investment to cover them, policy restrictions on the business opportunities of information providers should be reduced.¹³ This is the essence of the arguments advanced by telephone companies, although companies differ in their preferences for what restrictions are most critical to relax.¹⁴ Foreshadowing if not encouraging recent legislative and rulemaking debates, the NTIA had recommended relaxing restrictions on networks owning cable systems,¹⁵ common ownership of television stations and cable systems in the same market,¹⁶ national multiple ownership,¹⁷ and foreign ownership. Its own interests affected, the cable television industry contends that restrictions on cross-ownership should be relaxed only if the Congress acts to repeal state laws barring cable companies from providing telephone service.¹⁸

Greater freedom for telephone companies may involve unbundling or providing other service providers more direct access to basic network facilities and service elements. To that end, some have submitted proposals to the FCC and state regulatory authorities for unbundling their networks.¹⁹ The theory behind these proposals is that, by giving up their monopolies, regional telephone companies become deserving of entry into other markets, such as long-distance services. However, the technology of open data networking makes unbundling both easier to achieve and economically more compelling than it has been, although

most telephone companies may continue to resist ceding the control and revenue streams historically associated with bundled service provision (CSTB, 1994b). That is, the nature of the new technologies threatens to make traditional approaches to assessing infrastructure investments obsolete.

Noted Noam, ''Private carriers are now increasingly competing with common carriers, and conversely, common carriers will increase and compete with what private carriers do today. Those [situations] require all kinds of new forms of legal arrangements for dealing with traditional problems." Noam elaborated on the underlying economics:

In speculating on the need to change common carriage arrangements, Noam posited that most of "the policy goals behind common carriage, which are free flow of information, limited liability, reduction of transaction costs, and a whole bunch of other things … will be taken care of in a competitive system of private carriers. … The ones that are not necessarily going to be taken care of involve the free flow of information aspect. That would then suggest not so much making everybody a common carrier, but rather making sure that no carrier can discriminate

based on certain content criteria or some other long-term policy. This would be better than hanging on to the notion of common carriage per se."

Clearly, the debates over advancing the national information infrastructure underscore that development of nationwide networks will be very expensive; by one estimate, the physical facilities alone could consume $422 billion over 20 years.²⁰ In addition, the nation's reliance on the private sector for infrastructure investments demands that both public and private facilities be addressed in national policies (Egan and Wildman, 1992). Although volatile cost estimates reflect political considerations and business maneuvers, and although concerns have been raised about various risks, proponents argue that much of the investment would be made anyway,²¹ and that the payback to investors (let alone to society generally) will be worth it.²²

Apart from its controversial price-capping strategy—a relatively new approach to the regulation of information industries—the Cable Television Consumer Protection and Competition Act of 1992 demonstrates to members of the colloquium steering committee that Congress is often more amenable to retrospective "police" work than to forward-looking action on communications issues of the future. The chilling effect on cable-related investments and alliances ascribed to recent rate rollbacks is a cause for concern. And yet, even the Cable Act contained within it provisions that may have forward-looking impacts. For example, provisions relating to compatibility between cable service delivery and consumer electronics (TVs, VCRs) could serve to promote greater interoperability, notwithstanding industry concerns about specific mechanisms for doing so.

Noam elaborated on several points. The traditional telephone rate-subsidy system²³ will have to be restructured, he said, through some form of neutral surcharge and a voucher mechanism as a replacement for the present myriad pools, funds, and transfers. Through a principle of "third-party-neutral interconnection," a carrier could pick and choose its own customers, but it could not pick its customer's customers. This change will be brought on by increasing competition; providers will tend to compete for users who pay, and lower-revenue user groups will be left with fewer options and, as a result, with increasing prices. "These transitions will be painful. You'll hear from the losers, and [the transition] would require, therefore, leaders," Noam said. Competition will increase efficiency and hence reduce the magnitude of any subsidy, but it is not likely to decline to zero. Hence, we must reform the way it is being paid for, and reconcile it with competition. One idea, for example, is for every company providing telephone service to contribute in proportion to its customer base.²⁴

Gilder, on the other hand, argued that the need for subsidies will disappear altogether, because digital wireless technology will reduce cost disparities between rural and urban services. This view is not universally accepted, given the bandwidth limitations of wireless technology. But Gilder's optimism about available bandwidth leads him to anticipate the obsolescence of the entire regulatory and legislative structure that controls information services (Kelly, 1993). In its

auctions of spectrum for personal communications services (PCS), the FCC designed the auction so that even rural regions will have at least five major cellular and PCS providers. This approach should speed investment and lower prices.

The issue of how to assure universal service in a competitive marketplace also requires attention. As the information infrastructure evolves, people have begun to ask whether it will close or widen the gap between the rich and the poor. The Clinton administration has said it is "committed to developing a broad, modern concept of universal service—one that would emphasize giving all Americans who desire it easy, affordable access to advanced communications and information services, regardless of income, disability, or location" (IITF, 1993, p. 8). The progress during 1993-1994 of various new service trials, administration policy statements and explorations through the Information Infrastructure Task Force and the National Information Infrastructure Advisory Committee,²⁵ and legislative debates point to the difficulty of defining what it is that should be universally available. Mused Robert Johnson when interviewed,

It is difficult, for example, to identify what kinds of information may relate to different kinds of gaps in socioeconomic status; it is difficult to build a case for mandating simultaneous access to a broad range of new services when experimentation and exploration of what kinds of services will be most popular and how costs can be optimized are still at early stages.²⁶

Borrus, more optimistic than Noam about the future of common carriage, suggested that different types of carriers be assigned different responsibilities:

Regardless of how policymakers seek to guide or control information industries, the federal government can play a valuable role in promoting technological progress in other ways. In particular, colloquium participants stressed the effectiveness of procurement policies in improving government efficiency and diffusing new technologies throughout the economy.

Noting that the federal government is a major force in technology development,²⁷ Gilder suggested that government's role in the digital convergence should be that of an "aggressive, forward-looking consumer," purchasing the best possible technologies for interconnecting its supercomputer centers and other facilities. Such connections would maximize efficiency, according to Gilder, who has contended that the government typically "discovers a technology after its moment is passing" (Kelly, 1993). Gilder's comments echo other discussions about how the effectiveness of the federal investment in the Internet, whose backbone initially connected a variety of research and education facilities, contrasts with other federal efforts to promote specific technologies for government use, such as the Open Systems Interconnection protocol suite through the GOSIP federal information processing standard.

Borrus also has pointed to the influence of procurement policies. In assessing federal research and development (R&D) sponsorship in general, Borrus has contended that direct federal support historically has been less important to the success of commercial technology than have federal procurement and indirect support, because the latter approaches have tended to diffuse new technologies into widespread use.²⁸ On the other hand, the administration's National Performance Review (Clinton and Gore, 1993) and perennial efforts by adminstrations and the Congress to improve procurement policy and practice attest to the fact that procurement can be at best a blunt instrument. Also, the blossoming of the U.S. computer and communications enterprise provides evidence of the long-term fruits of federal R&D support overall (CSTB, 1995).

To further foster growth of the information infrastructure, Borrus suggested that policymakers consider how to stimulate experiments involving new technologies, configurations, and services, so that new applications and demand can emerge to support further evolution of the network portfolio. Similarly, Noam suggested that the government provide seed money for demonstration projects, as the National Science Foundation (NSF), the Advanced Research Projects Agency (ARPA), and other agencies have done. For example, NSF and ARPA contributed $15.8 million to the Gigabit Testbed program, which involved researchers from universities, national laboratories, supercomputer centers, and telephone and computer companies (CNRI, 1992). More generally, several colloquium participants identified the Internet as a working model for the type of experiment the federal government could support.²⁹ Experiments and demonstrations can help overcome a basic obstacle to network development: until a network is

established, its usefulness is difficult to prove, but a sponsor must pay for the project first. "So you run into the catch-22 where you can't develop demand until you're willing to pay for the network, but no one is willing to pay for the network until you've got some demand developed," Borrus explained.

At the same time, Borrus noted, oversight is needed to "permit a kind of a logical infrastructure to emerge which would emphasize standards for access interoperability between these various pieces of the network portfolio—levels of privacy, reliability, performance standards, and the like—but which wouldn't determine the actual implementation of any of the networks themselves." The National Cable Television Association (NCTA) has provided specific suggestions along those lines, arguing that local telephone companies should be required to "interconnect their facilities with those of more specialized providers on reasonable and nondiscriminatory terms" (NCTA, 1993a). NCTA also supports a federal role in standards setting for networks and customer equipment, to ensure, for example, that computers and telephones can send and receive cable signals and programming, as well as tax incentives to stimulate investment (NCTA, 1993a). The current fiscal climate makes it relative unlikely that tax incentives will be initiated. Moreover, whether and what kind of oversight is possible to achieve logical connectivity and so on is a challenge at the heart of public policy for advancing the information infrastructure: it is a challenge that many recognize but few can resolve to the satisfaction of the majority.

Commercial activities in the Internet arena have built on contributions to the technology base through federal research programs (such as the High Performance Computing and Communications Initiative) as well as development by private enterprise. However, competition, deregulation, and industry consolidation appear to have reduced telecommunications industry R&D from what it was in the heyday of AT&T's Bell Laboratories and in the early years of Bellcore. Thus Wildman has noted that U.S. telephone companies lag their foreign counterparts in R&D spending and patent activity (Egan and Wildman, 1992).³⁰ The combined forces of deficit-reduction pressures and a policy shift toward applications may constrain federally funded research in related technologies. Any slowdown in research would be particularly troubling now, because integrating diverse systems to evolve the nation's information infrastructure poses new technical challenges, some associated with lowering the costs of providing more general and flexible technology as compared to more specialized technology that is optimized for the delivery of a single service (CSTB, 1994b).

As noted in earlier chapters, a valuable part of the infrastructure is the content—the software and the programming. Thus, it is important for the government to support experiments focusing not only on networking hardware, but also on applications. Part of the appeal of the Internet, it should be noted, is that users develop their own "programming" and other information resources, from articles to databases, resources that can be located and interconnected increasingly easily

through the search and navigation tools (e.g., gopher, World Wide Web) that are themselves being developed by users (Berners-Lee et al., 1994).