Interoperability, Standards, and Security: Will the NII be
Based on Market Principles?
The United States stands on the verge of a quantum leap in the capabilities of its national information infrastructure (NII) 3. Widespread use of fiber optics is one element of that leap, as is the incredible advance in microprocessing, which has supported breakthroughs in digital compression. These, when applied to video, have made it possible to dramatically increase available communications bandwidth.
Just 5 short years ago, all-digital video was seen as impossible before 2005, if then 4. The amount of data or information required for full-motion video is orders of magnitude greater than that required for voice or nonvideo data transmissions. The data for video demands bandwidth beyond that required for voice and nonvideo data and requires effective compression techniques, to use the bandwidth efficiently or to meet some existing bandwidth limitation. In June 1990, General Instrument Corporation announced that it had solved this problem and would present an all-digital, high-definition television (HDTV) broadcast transmission system for consideration by the Federal Communications Commission and its Advisory Committee for a U.S. advanced television standard. One writer called this the technology equivalent of the "fall of the Berlin Wall" 5.
The conversion of video into a usable digital data stream is the third leg of the multimedia stool. The other two legs are data and voice telephony, the first being inherently digital, the latter having undergone a transition from analog to digital over many years. Carriage of video, voice, and data constitutes the foundation of the advanced broadband networks now being developed for both telephone and cable television networks.
Behind this simple concept, however, is a roiling pot of players and special interests, each trying to invade the other's turf or at least protect its own. Convergence is real because digitization breaks down the old barriers between service providers 6. It is increasingly hard to tell computer companies from video-equipment providers. It becomes increasingly more difficult to tell cable television operators from local exchange telephone companies. The growing community of Internet users adds to this mix, some of whom have an almost religious fervor about its uses and potential.
All of these players approach the information superhighway differently. Their perspectives and biases, at least initially, have been parochial and have caused them to overlook matters and issues that are important to other groups. The well-worn parable of the blind men and the elephant is apt here. The NII presents different characteristics and different issues, depending on how it is approached. When these converging players sit down with each other, they often find that they not only think or see things differently, they even tend to use a different vocabulary. What one set of players views as communication is mere babble to another.
Put another way, the convergence of technology is proceeding well in advance of the "cultural" convergence of the various segments of the industries. The resulting different perspectives influence the public policy debates that surround the NII. This paper addresses some of the issues of those debates, based on the following assertions:
These assertions may be controversial; many run counter to conventional wisdom, particularly that of the Washington, D.C., community. However, all are market-oriented. And they are fundamental to U.S. economic goals.
The NII: Not Synonymous with the Internet
Entertainment is the engine that has already pulled broadband networks into over 60 percent of American homes. Entertainment will drive the investment necessary to upgrade those networks as well. This is not a fact easily accepted by some Internet aficionados and some in the computer industry. Many in those groups join millions of other Americans who have a low opinion of much of current television programming. Many view the Internet as a liberator from TV's vast wasteland. But there are also other reasons why the role of entertainment is not more widely acknowledged by the computer world 7.
Computer businessmen are aware that today's dominant entertainment terminal, the television set, is an extremely cost-sensitive (and low-cost), relatively simple piece of electronics, geared to nonbusiness consumers, with a life of over 10 years. This is a long way from their preferred business model. The consumer electronics industry is dominated by foreign-owned companies; the computer industry is U.S. based. The television industry has relied on interlace scanning and sees it as important to keeping down the cost of its investment 8; the computer industry wants progressive scanning formats.
Likewise, those in the computer business look at other players and find them to be quite different from themselves. The cable television industry has only recently emerged from its "pioneer" phase and still revels in its "cowboy" image 9. Cable operators have an obsession with cost control, based on experience with mass consumer marketing. Where the computer industry has traditionally sold into the business community, cable operators have focused on residential customers. Regional Bell operating companies (RBOCs), in contrast to both, have previously been large, sleepy, and cumbersome bureaucracies, typical of utilities, and the antithesis of the computer industry 10.
The computer and Internet communities are not alone in their suspicion of the alien cultures that have suddenly entered their world. About 15 months ago, a high-level business manager 11 for a leading supplier to the cable television industry told his staff, "Somebody is going to have to explain this Internet to me!" He was reflecting not only his lack of knowledge of the phenomenon but also his exasperation at the whole range of new factors he had to consider as he developed his current core businesses.
So it is not surprising that computer business managers tend to gravitate toward an Internet model of the future NII. They will naturally have less enthusiasm for the role of entertainment than those who have worked in that field 12. But to recognize entertainment as the primary engine of the deployment of advanced broadband networks does not denigrate the role or the importance of the Internet. The NII is and should be about a lot more than just selling pay-per-view movies or making it possible for people to watch reruns of Roseanne or Baywatch. The advanced broadband networks that will share the task of serving as the backbone of the NII are about making video an integral part of all communications. The addition of video capability has major, positive implications for education, health, and business efficiency.
What entertainment can do is bring this broadband capability to every home and business. It can and will carry the major load of the investment needed to do that. When cable television operators begin to deploy digital decompression terminals, they will be putting into each user's home a level of computing power that is the equivalent of yesterday's mainframes. Far from detracting from or conflicting with the Internet, the broadband pipes of these networks will make Internet access via high-speed connections available to an increasingly wider range of Americans 13.
Achieving Open Systems
Open systems do not have to be, and perhaps should not be, nonproprietary; interoperability is not the only or even the major public good that must be served. For 2 years, the rallying cry has been, "We must take steps to ensure that the new technologies result in 'open and interoperable' networks." The General Accounting Office has proclaimed that "interoperability will be difficult to achieve" 14. Members of Congress have echoed this belief 15.
But is the situation really dire? Is there really a problem here? Is government intervention required? Thankfully, the conclusion is "no," because the cure could easily be worse than the disease.
Of course, interoperability is actually quite simple to achieve. All you have to do is standardize all elements of advanced broadband networks and the equipment that will use those networks. This can resolve all issues of compatibility, portability, and interoperability. Requiring compulsory licensing of any intellectual property that underlies those standards then provides unlimited competition in the supply of equipment for the networks. It is that simple, at least in concept. But even laying aside the political and legal barriers to such a program, achieving interoperability through central planning is a daunting, even overwhelming, technical task.
The view that interoperability is a major problem and that it might require such stultifying and intrusive steps rests on an outmoded view of communications capacity. In an environment of limited bandwidth, legitimate questions have been posed about the allocation of that scarce resource. When a cable television system carries 36 channels, issues about who gets "shelf space" in that system can take on major importance. Broadcasters demand "must-carry rules," because they are worried about retaining access to viewers. Programmers express concern that their products will receive less favorable treatment than those developed by the operator on whose system they seek to be carried. New services are concerned that they must go through an established multiple system operator to obtain the critical mass of viewers necessary for viability.
In an environment of bandwidth abundance 16, the incentives of network providers look quite different than they do in an environment of bandwidth scarcity. This is particularly true where the network provider faces competition. To pay for the network, providers have an incentive to include as many products and services as possible. They also have an incentive to make interconnection as easy as possible, and to develop consumer-friendly approaches to equipment design. For some purposes, network operators will continue to function in a broadcast mode. As channel capacity dramatically increases and fewer homes are served per node, each home or business served has the equivalent of one or more dedicated 6-MHz channels 17; this service, which resembles carriage more than broadcast, can be expected to develop as a result of consumer demand.
The incentives created by competition and broadband abundance will include incentives to achieve interoperability. Moreover, interfaces on these networks will remain open and do not have to be nonproprietary for this to occur. Network providers will avail themselves of the benefits of proprietary technologies but will, in their own interest, require that they be open to ensure competitive markets for equipment 18.
Just as the technology, by altering incentives, makes interoperability more likely or more easily achieved, so also is it inherently more conducive to interoperable solutions. Problems that existed in an analog environment are more easily solved in a digital environment.
During the early days of NII debate, Vice President Al Gore used a metaphor that was intended to justify a government role in setting ground rules for the NII. The vice president referred to the need for setting standards for interfaces on the NII and drew an analogy to the standardization of the gauges of connecting railroads 19. Analysis of this metaphor suggests a major distinction that tells us something important about the digital environment.
Railroad gauges, in fact, provide the perfect illustration of the digital environment's inherent ability to provide new answers to old problems. The chief difficulty created by the use of different gauges by connecting railroads was that, to move from one to the other, freight and passengers had to be transferred from one car to another car suitable for the different gauge. People and goods moving in interstate commerce could not be efficiently transported under such circumstances 20.
However, because of the essential nature of digital communications, data can be off-loaded, reloaded, and moved from one type of network or environment to another with comparative ease. Indeed, this flexibility is fundamental to the digital revolution, the foundation of convergence, and the basis of multimedia. In this
environment, the question of standards can often be replaced by another set of questions that relate not so much to the difficulty but rather to the cost of conversion. That places the matter in the hands of the market, not of regulators.
Moreover, in the digital world, breakthroughs continue. When addressing the issue of selecting the format for its high-definition television system, the Grand Alliance 21 discussed whether it should employ an interlace scanning format or a progressive scanning format. Upon investigation, it concluded that receivers could easily and economically convert the signals from one to the other. The result was that the system will support six formats and provide both interlace and progressive scanning.
A similar experience is occurring in the case of the modulation schemes used by different networks. Satellite and cable systems use different modulation. The Grand Alliance system currently in testing would use yet a third method for broadcast television. The development of multiple systems is not surprising, given the fact that network operators want to maximize their networks for their specific needs and use. There are actually good and sound reasons to tolerate the lack of interoperability that this could create. But there, it appears, potential issues may evaporate because CableLabs has indicated that it thinks dual or multiple modulation chips are feasible and economical 22. Thus, because of the inherent characteristics of digital technology, there is reasonable promise of avoiding inconsistency among networks.
Where interoperability is being addressed in the marketplace as a result of competition 23, incentives, and the technology's ability to support conversion, it avoids negative costs of standardization. That is indeed a good thing because standards are a mixed blessing.
Standards can and do inhibit innovation. The problem was outlined in a recent speech by GI's CEO, Dan Akerson:
The lesson of the personal computer success story is that the governmentshould not prescribetechnological standards in dynamic industries. Such standards freeze thecurrent level of technology inplace and they stifle the development of new technologies. When the governmentlets the marketplaceoperate, innovators innovate, competition flourishes, and consumers' choicesincrease. And, finally,equipment prices plummet. When the time is right, the technology will matureand the market will setstandards and insist on interoperability, the need for competitive pricing,and the availability ofcompatible equipment. 24
Where standards cannot be avoided, they should be narrowly focused. For that reason, interface standards are preferable to standardization of functionality, such as compression, transmission, and transport. In effect, interface standards at least restrict the "zone" where innovation can be stifled and provide some opportunity for innovation to develop around them 25.
There is yet another element to standardization that should give pause to those public officials who would apply it to these emerging industries. Standardization tends to commodize a product or technology. In their book Computer Wars, an insightful study of IBM and business strategies in the computer industry, Charles H. Ferguson and Charles R. Morris describe the effect of standardization:
Japanese companies, and Asian companies generally, have succeeded, by andlarge, by being superb commodity implementers within well-defined, stable, open, nonproprietary standardsstandards, that is, that are defined by regulatory agencies, other government bodies, industry standards-setting organizations, or very slow-moving incumbents, such as IBM has been in mainframes in recent years. Nonproprietary standard products, such as memory chips, printers, VCRs, CD players, or facsimile machines, are brutally competitive businesses, with high investment requirements and razor-thin margins.
But industries that are fast-moving, where standards are constantly evolving, and where the standards themselves are within the proprietary control of an individual company, are hostile environments for commodity implementers. And the computer industry in the1990s, under the technological impetus and creative impulse of the American start-ups, has been transmuting into just such an industry, shifting the ground out from under both the slow-moving Western giants and thecommodity manufacturing-oriented Japanese giants. 26…
The race over the next decade will be between the Japanese and American styles of industry development. The Japanese will drive to componentize and commoditize every sector of industry sotheir great monolithic and lean manufacturing skills can define the industry's future. American companies must keep accelerating the pace of technological change to avoid giving the Japanese a stable target to shoot at, and at the same time develop their own manufacturing skills to the point where low-end market share is not conceded too easily. 27
So the problem is not merely the stifling of innovation. Rather, it is that the very industries of critical importance to the U.S. economy are most susceptible to the negative effects of standardization. The digital revolution is uniquely the product of U.S. research and development. It represents a national asset, no less because it arrived on the heels of a period of "technology pessimism" in the late 1980s. American leadership was widely viewed as losing technology leadership for U.S. industries. Frequently, the problem was seen not as a failure of American invention but as a failure to capitalize on that invention. Many saw government-led industrial policy as the answer to regaining such preeminence.
Commoditizing the new technologies of the digital revolution will demonstrate that we have learned nothing from that earlier debate.
Who Should Set the Standards?
To the extent that standards are used, there is an issue about how they should be established. Here again, among the converging industries, differences in experience and practice create different perspectives. The broadcasting industry, for instance, lives in a world of government standardization, arising out of the practical problems created by potential interference among users of the radio spectrum 28. In addition, standards and regulations have pushed beyond this limited rationale to govern some features and functions of receivers. To an extent still being debated at the FCC, Congress has applied this form of standardization to the cable television industry as part of the cable television rate regulation legislation of 1992 29.
Similarly, the telephone industry is no stranger to standardization. In a unified Bell system, standardization was achieved within the company. But it has been the history of the telephone industry not only to live with standards but also to live with and even thrive under goverment regulation. Tell a telephone company executive that the government is contemplating standardizing some new product and the first question he asks is whether that isn't already regulated in his business; if it is, he dismisses it as a problem and moves on to the next issue.
The computer industry traditionally takes a different view of both standards and regulation than do the broadcast or telephone industries. Appropriately, it devotes significant energy and resources to remaining free of regulatory schemes and government-imposed requirements. There are industry standards, but they rise and fall very quickly and are constantly subject to attack and replacement. In the words of one broadcast official and part-time humorist: "The computer industry loves standards; it has a million of them."
The result of these differing attitudes, springing from different cultures and experiences, is a general policy approach that says the government should stay its hand and take a minimalist attitude toward standards and regulation, intervening only where there are overriding public interest concerns. However, this merely removes the debate to whether a given standard, regulation, or other proposal falls on one or the other side of that line.
The vagueness of this trigger for intervention and the fact that different biases are brought to its application by different groups of government officials invites politicizing of technology decisions. Thus, these issues become part of the political bazaar of Washington, a place ill-suited to making difficult technical choices. The private sector, not the government, should set standards; de jure standards, whether set by government or the private sector, are not preferable to de facto standards.
The computer industry is wise to oppose government-mandated standards 30. Even where congressional and regulatory officials can clearly formulate and enunciate their policy goals, their ability to penetrate to the next level of analysis and understand the technical ramifications of a given set of choices is severely limited. Government intervention is rich in the possibility for mistakes and for the operation of the "law of unintended consequences." And government almost always seeks compromise and consensus 31.
In recognition of these weaknesses in government intervention, many in the private sector, and some in government itself, have been advocating the position that these issues should not be addressed by government but should be left to private, standards-setting organizations. Decisions made in this fashion are preferable to decisions made by government officials. But there are good and valid reasons to ask whether substituting industry bureaucrats for government bureaucrats is, in the long run, a major improvement. There are valid reasons to ask whether this is a procedure that will yield the best results for the U.S. economy.
One problem is in the multiplicity of groups vying for the privilege of writing these standards. In the wake of the hype that accompanied the focus on an information superhighway, there are no fewer than 49 organizations or associations 32 claiming to have a role in the development of standards for digital compression or for the "set-top box" 33. Some of these groups are established standards organizations. Others are entrepreneurial in the Washington, D.C., sense: find an issue, organize around it, raise some money, and exercise influence. Some groups are national. Others are international. And in some you cannot be sure whether participants represent a U.S. or a foreign interest.
These different groups represent different perspectives, depending upon whether their members come exclusively or in part from the broadcast industry, the cable television industry, the consumer electronics industry, the telephone industry, the programming and production community, or the computer industry. This level of activity is one indicator of the reality of convergence; they are all converging on the "set-top box."
Another problem arises because private standards-setting efforts are as susceptible as is the government to attempts to use the process for some narrow end or to benefit some special interest. After all, losers love standards. The techniques vary but the result can be the same. Bureaucratic standards setting is slow moving, Ferguson and Morris complained 34. Pending standards can freeze the market and delay investment. Parties that are behind in the development process can use this delay to catch up to the industry leader. Frequently, those behind are numerous; the leader is, by definition, alone.
Like government, private standards-setting organizations generally rely on consensus 35. This kind of compromise yields a lowest-common-denominator approach 36. The fact that favoring consensus and not innovation can harm the economy is difficult for such groups and for the political process to comprehend; the effects are felt over time and diffused across the economy in general.
But the major problem arises because standards setting by bureaucratic bodies, whether public or private, is the antithesis of entrepreneurial development 37. The entrepreneur invents something, innovates. The product may or may not be technical. Then the entrepreneur "undertakes" 38 to bring together all of the things needed to take a product or service to market: financing, distribution, and so on. The entrepreneur's idea is then accepted or rejected in the marketplace.
By contrast, bureaucratic standards setting invents nothing. Its conclusions are never really tested in the marketplace (although, if they are too far afield, they might be rejected outright).
These standards-setting activities are a drain on resources for any company that participates 39. The smaller the company, the bigger the drain. Such activities require the diversion of valuable and scarce engineering resources. Thus, there is an inverse relationship between the size of the company and its ability to stay involved and protect its interests. Such private standards setting can work to the detriment of the entrepreneur to the extent that it prevents or limits his or her ability to obtain the fruits of the innovation. It can delay or even prevent the deployment of technologies 40. And, in addition to these factors, the entrepreneur can frequently end up outgunned and outmaneuvered. If he or she is the inventor, the other players seek to benefit from the invention.
In this environment, the entrepreneur gets little or no help from his government, which tends in any dispute to favor the side with more and larger players 41. What is more, government worries about de facto standards because they are messy: People might complain.
The most common argument against de facto standards is that "… we do not want to have another Beta/VHS situation!" How one feels about the Beta/VHS competition seems to be a matter of perspective, akin to the way that some people look at a glass of water. The "half-empty" crowd looks at Beta/VHS and calls it bad because it standed investment 42. But the ''half-full" crowd looks at the competition between Beta and VHS and concludes that it was a good thing because it resulted in improved VHS technology and ultimately provided consumers with more choices.
Rather than engaging in government standards-setting activities or in bureaucratic standards setting activities 43, what is needed is a new willingness (or a renewed willingness) to permit the market to work, to permit people to strike out and establish de facto standards, and to endure the uncertainty and even the temporary incompatibility that this sometimes creates. In the words of Joseph Shumpeter, "Let the gale of creative destruction blow through this industry."
Unless Security and Privacy are Protected, the NII Will not Reach itsFull Potential as a Platform for Electronic Commerce
The deployment of advanced broadband systems will depend upon private investment.44 The success of these systems will depend upon finding applications that will support that investment.
Deployment requires two conditions:
These are the forces that drive a system of dynamic competition in this rapidly evolving digital world. These forces need to be encouraged and protected. This can occur only if there is (1) a recognition that security for protecting intellectual property is a critical element, and (2) a recognition that security systems must be renewed, or evolve, through continued technological innovation if they are to keep pace with those who would seek to violate property rights.
Based on experience over many years, there are some fundamental facts and principles that should drive all policy:
Security for and protection of intellectual property are fundamental needs for the success of advanced broadband networks, but no less important is the ability of those networks to protect personal privacythat is, the privacy of personal information. If advanced networks are to become "platforms for electronic commerce," they will need to support a wide variety of applications, including delivery of electronic products, home shopping home banking, and medical assistance. The willingness of citizens and consumers to employ these networks for these purposes will depend upon their sense of confidence that their personal information is secure from prying eyes.
Proposals to require that equipment on these advanced broadband networks (and current cable and telephone networks, as well) be made available at the retail level 45 are an example of the way that regulating one aspect of a network can distort the operation of all others. In this case, the negative implications are for security. Consumer electronics retailers would, not surprisingly, make retail availability of equipment the central tenet of all policy. Naturally, proponents of this point of view ignore other problems that occur when this happens, such as freezing technologies or interfaces or creating problems for maintaining security. They would turn technology decisions into political decisions.
These proposals to mandate retail sale of consumer equipment used with the network are supported by dubious analogies and assertions. One such is the attempt to apply to video distribution and developing broadband networks the model currently used for voice telephone service, whereby an interface specification was developed and used to separate network functions from the functions performed by "consumer premises equipment" 46. The telephone model is viewed as a way to create competitive markets for equipment used to supply broadband services 47.
A major problem is that these proposals ignore the fact that voice telephone service requires only access control; video providers have traditionally (and for good reason, given the problems of theft of service) insisted on scrambling as well. Nevertheless, there are indications that the markets will develop such separation, and probably more quickly, without government interference 48. Another claim made by proponents of this type of government intervention is that new "smart-card" technologies are available. They cite draft (not even finalized) Electronics Industry Association standards for such security systems 49 and seem oblivious to the contentiousness regarding the adequacy of such standards and what features are needed to bring them up to acceptable levels of security 50.
The principle that underlies these proposals is that the government can and should determine how much security is adequate and that that determination will bind owners of intellectual property and network providers. That principle is unacceptable 51.
Owners of intellectual property and those acting for them should have the right to determine the level and method of security appropriate for their needs. These decisions should be made on the basis of business needs and realities, not by government.
As network providers experiment with different types of broadband networks 52, some of which have different implications for security requirements, the government should let these technologies work themselves out in the marketplace, free from the distortions that could be created when the government picks out one factor 53 among the many that must be considered, thereby affecting all other choices 54.
The assertions set forth here have in common a faith in and a reliance on market forces. The role of entertainment in driving deployment of advanced broadband networks and the need for security for those networks are related. Given adequate government recognition of the rights of property, and given government restraint from imposing other regulations that undermine those rights, these networks can develop. This will not happen instantaneously but will proceed incrementally over the next decade. The end result will be the extention of new broadband capabilities to all parts of the country.
Interoperability issues would also benefit from an understanding of the way markets are working to address problems far more rapidly than government or even private-sector standards organizations can accomplish. Market resolution recognizes the importance of incentives that motivate entrepreneurs. Failure to understand these markets and these incentives can have negative consequences for the economy and for international competitiveness. Conversely, if the government refrains from micromanagement and can successfully provide sound macroeconomic policy, the revolution that is the information age could bring a period of progress on the scale of the industrial revolution.
1. Quincy Rodgers manages government affairs for General Instrument Corporation (GI). A graduate of Yale College and the Yale Law School, he served as Executive Director, Domestic Council Committee on the Right of Privacy during the Ford Administration. He has also practiced law in New York City and Washington, D.C., and has served on the staff of the Senate Judiciary Committee.
2. GI is a world leader in broadband transmission, distribution, and access control technologies for cable, satellite, and terrestrial broadcasting applications. In 1990, GI was the first to propose and, in 1991, the first to deliver an all-digital HDTV system for testing in accordance with FCC procedures. GI has delivered all-digital compression systems for broadcast, satellite, and cable systems.
3. The national information infrastructure has been recognized as a separate subject for policy attention for at least 20 years. See National Information Policy, Report to the President of the United States, submitted by the Staff of the Domestic Council Committee on the Right of Privacy, Honorable Nelson A. Rockefeller, Chairman, National Commission on Libraries and Information Science, Washington, D.C., 1976.
4. This was the conclusion of at least one respected and generally knowledgeable business leader in the entertainment/communications industry. Japanese and European HDTV systems were apparently based on this false premise. The Japanese analog MUSE HDTV system was outmoded even as it was launched. The implications of this experience for government-led "industrial policy" are beyond the scope of this paper. See Cynthia A. Beltz, High Tech Maneuvers, Industrial Policy Lessons of HDTV, AEI Press, Washington, D.C., 1991.
5. Mike Mills, "A Digital Breakthrough," Congressional Quarterly, January 15, 1994, p. 66.
6. Computer Science and Telecommunications Board, National Research Council, Keeping the U.S. Computer and Communications Industry Competitive: Convergence of Computing, Communications, and Entertainment, National Academy Press, Washington, D.C., 1995.
7. Of course, entertainment includes games, and the computer community is very much a part of that product line.
8. Some television executives and engineers believe that interlace provides a better picture than does progressive, although admittedly not for text.
9. "Cable-TV Industry Loudly Likens Itself to Wild, Wild West," Wall Street Journal, May 10, 1995, p. B2.
10. Former FCC Commissioner Ervin Duggan, now president of PBS, captured the cultural diversity of at least two players in the converging arena when, on a panel at a Cable Television Show, he said that, despite their size and cash, telephone companies posed no danger to the current Hollywood and programming communities because telephone executives "still sleep in their pajamas." Bell executives may yet prove him wrong as they make major investments in creative material. Are the Bells going Hollywood? See "Staid Phone Giants Try Marriage to Hollywood," Wall Street Journal, May 24, 1995, p. B1.
11. Who will remain anonymous.
12. Following are some important differences between the computer and the entertainment worlds: computer is point-to-point communications, carriage, password-oriented access control where most messages have little intrinsic value; entertainment is point-to-multipoint, broadcast, and scrambled on broadband, and the material transmitted has great intrinsic value.
13. "Datacom Emerges As Next Wave," Multichannel News, May 15, 1995, p. 1.
14. Information Superhighway, An Overview of Technology Challenges, Report to the Congress, U.S. General Accounting Office, Washington, D.C., January 1995, p. 31.
15. H.R. 3626 (103rd Congress, 2nd Session), Section 405; S. 710 (104th Congress, 1st Session).
16. George Gilder, "When Bandwidth Is Free," Wired, September/October 1993, p. 38; George Gilder, "Telecosm: The Bandwidth Tidal Wave, Forbes ASAP, Supplement, December 5, 1984, p. 163.
17. Speech by Daniel F. Akerson, Chairman & CEO, General Instrument Corporation, at Lehman Bros. Conference, San Francisco, California, April 28, 1994, Slide 5:
… By compressing 6 digital channels into the space of 1 analog channel, you can get 660 channels per today's node of 500 homesor one personalized channel for each home. By further compressing at 10-to-1 ratios, using compression technologies such as GI's digicipher, a broadband system can provide 7 personalized channels per home. At this level of service, the consumer has the ability to be online from multiple TVs and PCs in the home.
18. General Instrument Corporation's digital television technologies are licensed under terms of agreements that it has with cable television operators. GI has licensed three U.S.-owned manufacturers (Scientific Atlanta, Zenith, Hewlett-Packard) to build its complete systems, including access control (security). GI has also indicated that it is prepared to more widely license its digital decompression technology. Parenthetically, the case seems strong for more narrow licensing of security technology than for other technologies. Although these licensing arrangements were established, in the first instance, as a result of vendor agreements, there are also strong economic incentives for vendors to license their technologies in order to drive their acceptance in the market. Charles H. Ferguson and Charles R. Morris, Computer Wars, Time Books, Random House, New York, 1993, p. 143 (hereinafter Computer Wars). In addition, the complexity of digital communications systems requires that someone perform the functions of disseminating information, answering questions relating to the system, and conducting compliance and interoperability testing. GI's licensing program addresses these needs.
19. Remarks prepared for delivery by Vice President Al Gore, Royce Hall, UCLA, Los Angeles, California, January 11, 1994.
20. Imperial Russia turned this shortcoming into an asset by maintaining gauges different from those of its Western neighbors as an additional deterrent to invasion. Lessons for the NII to be drawn from the experience of the U.S. railroads do exist: "Between 1865 and 1873, 35,000 miles of track were laid, a figure that exceeded the entire rail network of 1865, and inspired a boom in coal and pig iron production … and the rapid spread of the new Bessemer process for making steel. Railroads opened vast new areas to commercial farming.… Their voracious appetite for funds absorbed much of the nation's investment capital … contributed to the development of banking … and facilitated the further concentration of the nation's capital market.…" Eric Foner, Reconstruction: America's Unfinished Revolution, 1863–1877, Harper & Row, New York, 1988, p. 461.
21. The Grand Alliance is made up of the companies and organizations that were proponents of HDTV systems as part of the FCC's Advisory Committee on Advanced Television Service (ACATS) program for developing an advanced television standard. Participants were AT&T, General Instrument, MIT, Philips, Sarnoff, Thomson, and Zenith. AT&T and Zenith jointly submitted a progressive scan system for testing. Philips, Thomson, and Sarnoff jointly submitted an interlace system. GI and MIT jointly submitted two systems: one progressive and one interlace.
22. "Universal Demodulator Handles QAM, VSB and NTSC," Communications Technology, December, 1994, p. 100.
23. An interesting paper prepared for the American Enterprise Institute calls into question much of the conventional wisdom about interoperability. Milton Mueller has taken a historian's look at the development of the telephone companies in the United States and examined the effect that interconnection requirements had on the rollout of telephone service in the late nineteenth and early twentieth centuries. Mueller maintains that the absence of interconnection requirements was an important element of the rapid penetration of telephone service in the United States. The inability of competing telephone systems to interconnect with each other, according to Mueller, forced competition between the Bells and independent telephone companies to focus on providing access. In the absence of interconnection, the desire of potential customers to have access to telephone service and to communicate, not with everyone or anyone, but with specific groups of other users, became the central focus of competition. This resulted in more rapid penetration of service, driven by suppliers of that service. It became a race between the Bells and other providers to see who could sign up the most customers and most rapidly expand the network. This thesis may have implications for public policy and the rollout of new, digital services. Milton Mueller, "Universal Service: Competition, Interconnection and Monopoly in the Making of the American Telephone System," prepared for the American Enterprise Institute for Public Policy Research, working paper presented March 31, 1995.
24. Remarks by Daniel F. Akerson, Chairman and Chief Executive Officer, General Instrument Corporation, at the Washington Metropolitan Cable Club Luncheon, Washington, D.C., April 11, 1995. See also Peter K. Pitsch and David C. Murray, "A New Vision for Digital Telecommunications," A Briefing Paper, No. 171, The Competitiveness Center of the Hudson Institute, Indianapolis, Indiana, December 1994.
In addition, see Stanley M. Besen and Leland L. Johnson, "Compatibility Standards, Competition and Innovation in the Broadcasting Industry," Rand Corporation, November 1986, p. 135. According to Besen and Johnson, "The government should refrain from attempting to mandate or evaluate standards when the technologies themselves are subject to rapid change. A major reason for the Commission's difficulty in establishing the first color television standard was the fact that competing technologies were undergoing rapid change even during the Commission's deliberations. It is
only after the technologies have 'settled down' that government action is most likely to be fruitful, as illustrated in the TV stereo case."
25. Even interface standards have the potential to inhibit development. The Consumer Electronics Group (CEG) of the Electronics Industry Association and the National Cable Television Association (NCTA) have been struggling over the development of a decoder interface specification to be adopted by the FCC. This standard is seen by the Commission as a solution to issues of compatibility between televisions/VCRs and cable television systems. The proceeding is a response to the Cable Act of 1992 (Cable Television Consumer Protection and Competition Act of 1992, P.L. No. 102-385, 16 Stat. 1460 (1992), Section 17). NCTA and some cable industry suppliers are seeking an interface that can utilize a more extensive command set and more connector pins, maintaining that this will be needed to support future services, including QWERTY keyboards, an air mouse, and other yet-to-be-developed but envisioned capabilities. Alternatively, the cable side has argued that the interface should permit the "pass through" of consumer remote control commands so that the consumer can utilize all features of the network. CEG is resisting these requests, arguing, inter alia, potential interference with the features its members provide in consumer electronic equipment. Both sides have made extensive filings with the Commission in an ever growing record (In the Matter of Implementation of Section 17 of the Cable Television Consumer Protection and Competition Act of 1992; Compatibility Between Cable Systems and Consumer Electronics Equipment, FCC ET Docket No. 93-7). Recently, elements of the computer industry have entered the proceeding, voicing concerns over the adequacy of the proposed interface and at least one feature previously agreed to by CEG and NCTA. A Silicon Valley company has mounted a campaign against the proposed solution, charging it can become a bottleneck.
26. Computer Wars, pp. 113–14; see also fn. 17. During the early days of the Clinton Administration, this book was reportedly widely read in the White House. Current public policy debates provide a rich set of opportunities for the Administration to heed its lessons.
27. Computer Wars, p. 221; see also fn. 17.
28. Even this long-standing and "hallowed" justification for government regulation is being called into question. See Adam D. Thierer, "A Policy Maker's Guide to Deregulating Telecommunications, Part 1: The Open Access Solution," Heritage Talking Points, The Heritage Foundation, December 13, 1994; Paul Baren, ''Visions of the 21st Century Communications: Is the Shortage of Radio Spectrum for Broadband Networks of the Future a Self-Made Problem?," keynote talk transcript, 8th Annual Conference on Next Generation Networks, Washington, D.C., November 9, 1994.
29. Cable Television Consumer Protection and Competition Act of 1992, P.L. No. 102-385, 16 Stat. 1460 (1992), Section 17.
30. Computer Systems Policy Project, "Perspectives on the National Information Infrastructure: Ensuring Interoperability," white paper, February 1994; Alliance to Promote Software Innovation and the Business Software Alliance, "The Information Marketplace: The Perspective of the Software and Computer Industry," special focus paper, Spring 1995.
31. Government standards, like all standards, need to take into account the international circumstances in which the standard will operate. The U.S. market remains the world's most attractive, thus giving a U.S. standard additional impetus. In the digital video arena, however, there are signs that Europe or Japan or both may gang up and try to protect their domestic industries by promulgating a standard different from that used in the United States. At a recent meeting in Melbourne of DAVIC, a group working on an international cable standard for digital video, the Japanese and European representatives joined forces in opposing the position of most U.S. representatives. It is too soon to tell whether this kind of activity will ultimately undermine the leadership position that the United States has held as the home of digital development.
32. This number was used by FCC Chairman Reed Hundt in a speech to the National Cable Television Association Annual Cable Show, May 9, 1995.
33. The origin of this term is unclear. Some early commentator made the statement that the square foot on the top of the television set was the most valuable piece of real estate on the information superhighway. This immediately caught the attention of at least one member of Congress who worried that it could become a bottleneck. It also brought the attention of many companies that announced that they planned to make "set-top boxes," long before they had thought through what functions the set-top box would perform or what applications consumers wanted. The term "set-top box" is almost certainly derived from the model of current cable television converters. The convenience that the use of the term provides is more than offset by the way it misdirects focus from the real issue, which is where different functionality is to be performed in a network.
34. Computer Wars; see also fn. 17.
35. Sometimes these efforts operate on a principle of unanimity. At other times, minority views are outvoted and rejected. Some groups are incredibly undemocratic; prudence requires not identifying them here, now. The United
States government should acquaint itself with the characteristics of such organizations prior to embracing or affirming their policies.
36. "… The bureaucratic solution is most frequently 'one size fits all.'" George Gilder, speech at the Virtual Democracy Conference, Progress and Freedom Foundation, January 10, 1995. In remarks also applicable to this subject, Gilder said that the answer to NII development lies not so much in competition as in diversity.
37. "… Most entrepreneurs who make breakthroughs are almost by definition psychologically uninterested in working in a dull, slow process in which most of your time is spent listening to people who don't know as much as you do." Speaker of the House of Representatives Newt Gingrich, speech at the Virtual Democracy Conference, Progress and Freedom Foundation, January 10, 1995.
38. Translation provided by Speaker Gingrich, in remarks to the Computer Science & Telecommunications Board, Washington, D.C., May 15, 1995.
39. From a slide offered by CableLabs at a recent Cable/IT Convergence Forum (and, one hopes, intended only to wake up the participants):
40. General Instrument Corporation's DigiCable product, which will allow cable systems to offer the equivalent of 350 to 500 channels, is approximately 1 year late to market. Six months of this is attributable to technical problems that the company and its vendors encountered; these digital consumer decoders are the most sophisticated pieces of consumer electronics hardware ever deployed. But 6 months of the delay was caused by a redesign to meet the specifications under development by the Moving Picture Experts Group for MPEG-2, even though MPEG-2 does not constitute a complete transmission system while DigiCable does. "Digital Compression: The Holy Grail," Cablevision Magazine, March 6, 1995, p. 22.
41. The tendency of politicians to count noses can work against important national resources. As foreign companies locate plants in the United States, they bring valuable employment to Americans. However, jobs are not the only contribution that technology makes to national wealth. In particular, U.S. ownership of and benefits from research and development activity are an important component of economic investment. Among senior advisers to the Clinton Administration, the relative merits of these factors have been the source of debate. See Robert Reich, "Who Is Us?" Harvard Business Review, January–February 1990, p. 53; and Laura D'Andrea Tyson, "Strong U.S. Firms Key to Healthy World Economy," Los Angeles Times, December 9, 1990, p. D2. To oversimplify, Reich thinks that foreign firms that provide jobs should be considered American; Tyson thinks that they must do R&D in the United States to qualify. Neither expresses concern about foreign ownership in the context of national and international standards setting; European governments and European multinationals, in particular, have demonstrated an all too frequent tendency to use standards to protect domestic industries.
42. Providing a government entitlement or backward compatibility protection for early adapters who purchased Beta machines for $1,000 and more would not seem to be a major national priority; presumably they got something for their early purchase, namely to be the "first ones on their block.…" In any case, the economy provides winners and losers. This author maintains an extensive collection of Beta tapes and offers to buy used Beta machines in good working condition for $25 to $35 and $10 to $15 for those in need of reasonable repair. Thus, a secondary market, of sorts, has already developed.
43. Or, worst of all, joint government/private-sector standards-setting activities!
44. Much of the material in this section is drawn from testimony of Richard S. Friedland, president and chief operating officer of General Instrument Corporation, before the NII Security Issues Forum, July 15, 1994.
45. H.R. 1555, Sec. 203 (104th Congress, 1st Session); H.R. 1275 & S. 664 (104th Congress, 1st Session).
46. The telephone specification was well established before regulatory rules and separation were imposed. Moreover, the installed base of telephones was huge, covering over 90 percent of the population. The current installed base of digital video equipment is minuscule by comparison and is exclusively one-way satellite. The technical complexity of voice telephone is orders of magnitude less than that of video networks. There is simply no comparison, a fact that did not inhibit members of Congress who showed up at a press conference announcing this legislation with antique, black, rotary-dial telephones.
47. The market for such equipment is currently highly competitive and has numerous providers, including General Instrument, Scientific Atlanta, Panasonic, Zenith, and Philips, among others. In anticipation of its potential
growth from the introduction of digital transmission and the advent of telephone company video services, others such as Hewlett-Packard, Thomson, Sun, and Sony have said they would enter. Today, telephones, like most consumer electronics, are produced by offshore companies, although many telephones carry the Bell label.
Retail sale proposals are advocated as consumer friendly and, as such, are supported by the organized public interest consumer groups. Retailers do not propose to cut off the right of network providers to also provide equipment, by sale or lease, but would prevent it from being bundled. Techniques such as those used so successfully by cellular telephone companies, whereby deep discounts for equipment allow more people to take advantage of service, would be barred. There are legitimate questions as to how "consumer friendly" it is to deprive consumers of the option of low-cost entry to these new services. Likewise, lease of equipment is likely to be an important component of the introduction of new services. It is one thing to buy a telephone for $15 to $100 from Radio Shack, knowing that the chances of its becoming obsolete are minimal. Indeed, it will almost certainly break before then. It is another thing to expect consumers to rush down to Circuit City to buy a $300 to $500 digital entertainment terminal and bear the risk of obsolescence and perhaps the risk that it will not work properly with the network.
48. Bell Atlantic, Nynex, and PacTel have issued an RFP for equipment, including separation of a network integration module (NIM) and a digital entertainment terminal (DET). The NIM would include network functions, including access control and decryption. The DET would be available for retail sale. See Chris Nolan, "The Telcos' Set-tops," Cablevision, April 3, 1995, p. 56. However, earlier this year, AT&T and VLSI Technologies announced a program that runs counter to the thrust of separation. See "AT&T, VLSI to Embed Security into Set Top Chips," Broadcasting & Cable Magazine, February 6, 1995, p. 36. AT&T customers, like GI customers, who are developing products, tend to focus, laser-like, on maximizing security. It is doubtful how many are aware that some in government have different plans for them.
49. Congressional Record, 104th Congress, 1st Session, April 4, 1995, p. S5143.
50. Although it is generally acknowledged that smart card technologies will be easier to develop and more secure in a digital environment, the experience in the analog environment is that security has been badly compromised. Satellite systems are badly compromised in Europe. See "Sky War Over 'Smart' Pirates," Sunday Mail, Financial Section, United Kingdom, October 9, 1994.
51. Interesting questions would be posed by such a government determination: What happens if nontechnologists in the Congress and or in the bureaucracies (or even qualified technologists, for that matter) turn out to be wrong and the technology is defeated by theft-of-service pirates? Since the government has chosen this technology, would or should the government be liable for the losses incurred as a result of the defeat of this technology? If the technology were defeated, could retailers be liable for losses incurred thereby? Would that liability require a showing of negligence, such as the failure to maintain secure warehouses for the equipment? Given that maintaining the security of these systems is a matter of concern to the government (the success of our communications networks and the investment in those networks depend upon their ability to maintain security and privacy), would it be appropriate that punitive damages be available in the case of a break against a party, including a retailer, whose negligence caused a break? Should tort reform legislation contain an exception for punitive damages in such cases? Could a manufacturer refuse to make products available to a retailer who failed to sign an agreement in which the retailer promised to maintain adequate security procedures or an agreement in which the retailer promised to indemnify against losses from a breach of security?
52. Hybrid fiber coaxial cable systems (HFC) seem the preferred model for cable television operators and some telcos. Other telcos are indicating a preference for switched digital networks (SDN). Ultimate choices will be worked out in the marketplace over time, as network providers experiment with different delivery models. Cost will be a fundamental issue.
53. For example, retail sale, or even including security considerations advanced without regard to cost or convenience.
54. Retail sale has implications for other values as well, such as innovation. To the extent retail sale is used to justify standardization, it can negatively affect continued development.