The world is shrinking in size, thanks to the ongoing information revolution. Most parts of the world are already connected by telephones. We can call almost any part of the world from almost any place. We take it for granted. Information service(encompassing voice, video and data) is on the brink of a similar phenomenon. There are already islands of information networks growing rapidly and merging with other growing networks. What stands between the ultimate connectivity and us today is the connection of these disparate networks and adoption and proliferation of standards. Various types of advanced communication systems have evolved in US, Europe and Japan to cater to various customer needs. Today, we are not only faced with the choice of various communication systems in a buying decision but also the problem of interconnecting existing communication systems in a seamless fashion. The revolution is not necessarily limited to networks, computer programs can now read files produced by a competitors program, for example Microsoft Word can read Word Perfect documents. In essence the revolution is about different hardware and software working transparently to increase ease of use and efficiency.
The Personal Computer hit the market in the late seventies and early eighties. With a choice of Apple, Commodore, IBM and other proprietary systems it was not clear to the users which one would become the standard. Over the last decade the IBM PC won over the rest as a de facto standard. The emergence of the IBM PC as the standard was not related to the underlying technology, but to the open architecture and the marketing prowess of IBM. The boom in the PC market conferred a monopolistic status on Microsoft and Intel. More than seventy-five percent of the personal computers in the world today run Microsoft's operating system and Intel has a similar share in the microprocessor market. Both Microsoft and Intel have staunchly defended their monopolistic status by protecting themselves with an array of copyrights and patents. Companies like AMD and Cyrix trying to enter the microprocessor market have faced hostile lawsuits from Intel. Microsoft has been accused of unfair trade practices like exclusionary and anti competitive OEM licensing and is currently being investigated for antitrust.
On the other hand telecommunication companies like AT&T had been extremely open. This was in part because of government regulation of the AT&T monopoly and in part due to the obvious network externalities flowing from a universal system.
The debate today is about the interfaces between software and between networks. Computer and Telecommunication technologies are converging into this grand infrastructure called the information superhighway. There are conflicting viewpoints between the two industries about the required degree of openness of the interfaces. Also, given the history of de facto monopolies of the past decade, many smaller companies and government agencies do not want a repeat. Most organizations do not want this revolution to spawn new monopolies or offer companies like Microsoft, Intel and IBM any additional advantage to sustain their monopolistic power. The debate about standards and interoperability is set in this backdrop of diverse beliefs and potentially huge monetary incentives.
Interoperability has been defined as:
"the ability of two or more systems (computers, communications devices, databases, networks or other information technologies) to interact with one another and exchange data according to a prescribed method in order to achieve predictable results" [2]
Al Gore was elected as the Vice President in 1992. Since Gore is a champion of the information infrastructure, Washington began to focus more attention on the information infrastructure. The Clinton administration in 1993 issued "The National Information Infrastructure: Agenda for Action," which sketched out its basic vision for the National Information Infrastructure (NII). One of the goals of the Administration was to "promote seamless, interactive, user-driven operation of the NII."[6] The Administration recognized that this goal could be accomplished only through interoperability. The Agenda for Action established the National Information Infrastructure Task Force (NIITF) to study among other things standards for interoperability.
While most organizations agree on the need for interoperable interface standards, the major disagreement seems to be about how these standards should be set and regulated. While there are organizations and people who believe that standards should be set by the government, there are companies at the other end of the spectrum who believe that such standards should evolve unregulated out of a free market. Three different approaches to standards have been proposed:
In this paper we examine the merits of the arguments of each of the three camps.
This approach is mainly advocated by NIST. NIST's premise is that market failures will prevent the achievement of optimal level of interoperability and government intervention is necessary to prevent this failure and maximize the aggregate economic benefit to society. NIST's arguments in support of this premise are given below.
Economic issues related to interoperability center around finding a balance between the need to minimize chaos and inefficiency caused by adopting multiple technical solutions and the need to maintain incentives for competition and innovation.[1] NIST argues that unless one solution gains a dominant position in the marketplace in a timely fashion, the potential for innovation and economic growth surrounding the technology is hampered. Perfect interoperability is not necessarily optimal for maximizing aggregate economic benefits in a given economy. Technical solutions that provide less than perfect interoperability may actually generate greater economic benefits if adopted at the proper time.
NIST believes that only by intervention is it possible to bring together industries, which have traditionally operated in different environments, to create the National Information Infrastructure. There is commonality inherent in the concept of interoperability which makes it infrastructural in nature; it is a feature which enables the development of elements of technology which are jointly used by competing firms. This requires a certain amount of cooperation, coordination and information sharing which is sometimes not promoted by market forces. In fact there might be disincentives for individual firms to be efficient resource allocation which maximizes aggregate economic benefits, as opposed to benefits accruing to the individual firms. In addition, it is becoming increasingly apparent that interoperability may require convergence of various paradigms for technical development and market competition from disparate and sometimes seemingly incompatible industries. The telecommunication industry for example, until recently was used to operating as a highly regulated, semi-monopolistic industry, and the computer industry on the other hand has enjoyed a more competitive market model with minimal regulation.
NIST feels that network externality is another important reason for government intervention. There are aggregate benefits to society at large generated due to enhanced interoperability that are beyond those captured by competing firms. In other words externalities are present. This means that the total economic benefit derived from the infrastructure increases more than linearly with the number of agents connected to the network. The aggregate benefit accrues up to a certain critical mass. Beyond that level, there is no additional benefit to be gained by increased activity. In fact, increased activity may produce additional costs which offset some of the benefit gained. For example greater aggregate benefits will be accrued with increasing degrees of interoperability. Those benefits accrue until a critical degree of interoperability is reached. At this degree of interoperability, the greatest benefit to the economy at large can be achieved. Beyond this level, aggregate benefits are offset as greater interoperability may result in reduced market competition and/or engender offsetting costs.
NIST does not take into account the fact that the computer industry has in the past shown a tendency to gravitate towards a de facto standard. The standards for PC and a lot of other software have evolved without government intervention. The computer industry has been growing at an increasing rate and has not shown any signs of failure. The Internet is an example of a highly flexible network which has evolved to connect heterogeneous systems. Though the Internet was initiated by a government agency, it has largely grown on its own without any government intervention. On the other hand government intervention has probably been responsible for making telephone a universal service. It was regulations and subsidies that enabled the remote rural areas to have access to telephone facilities. One of the problems with the government agencies setting standards is that technologies are changing and evolving very rapidly and government agencies are slow; by the time the standards are set it might be too late anyway because market might have embraced a proprietary standard. The other possible problem is the government setting standards prematurely, which stifles innovation and compromises U.S. competitiveness in the global market.
This camp believes that non copyrightable interface specifications developed by free market mechanisms will promote the delicate balance of competing public policy considerations embodied in the Copyright Act. The balance here is between the desire to serve public interest in the wide dissemination of information and the desire to provide incentives to authors to create new works. Protection of content will play an important role in the development of NII. At the same time, protection is not the only important consideration; an equally important consideration is promoting interoperability. Overly broad intellectual property protection for standard interface specification in the NII could inhibit such interoperability. This viewpoint is mainly supported by ACIS and companies like SUN Microsystems. Their main arguments are given below.
ACIS believes that the U.S. government (particularly the antitrust enforcement agencies) should specifically endorse the Second Circuit's seminal opinion in Computer Associates International Inc. Vs Altai, Inc., 982 F.2nd 204 (2d Cir. 1988).[3] The Second Circuit's abstraction -filtration-comparison test for copyright infringement, adopted in Computer Associates has been applied by Federal Courts of Appeal in other circuits. This ruling embodies legal principles absolutely critical to innovators seeking to develop hardware and software that is interoperable with systems that have, for either technical or market reasons, become standards. To extend copyright protection to interface specifications that become the standardized rules of interconnection in the NII would hinder interoperability, impede competition and innovation, and serve as a critical impediment to the implementation of the NII and the principles advanced by the Administration. A balance between access to these standards and protection of the implementation of these standards in program code is required and Computer Associates strikes the proper balance, at least with respect to copyright law.
ACIS also supports the Sega and Nintendo decisions with respect to the permissibility of disassembly. Software owners have the right to make interim copies of a work in order to separate protected and unprotected elements of a work which is essential to interoperability. ACIS argues that such fair uses for the purposes of reverse engineering have been endorsed by at least two federal courts of appeals.
ACIS believes that the Administration should urge Congress to adopt Section 405 of H.R. 3626, which requires the Federal Communications Commission to study whether critical NII interfaces should be open and accessible to developers of interoperable products.[4] The legislation advocates openness and accessibility of interface specifications which ACIS believes is necessary to "ensure diversity and thwart anti competitive practices" in the NII.
ACIS feels that the Administration should support the amendment of the Patent Act to improve the quality of patents issued, especially pre-grant publication of applications. As a consequence of the Computer Associates, Sega and Nintendo decisions, vendors have begun to seek U.S. patent protection for their interface specifications. Because of the PTO's lack of an adequate prior art database, its shortage of experienced examiners, and other deficiencies in the process, ACIS fears that the PTO will grant patents to interface-related inventions which do not meet the criteria of novelty and non-obviousness. ACIS also supports pre-grant publication of patent applications. This, they believe, will enable third parties to bring any relevant prior art to the examiner's attention and thus ensure that interoperability patents meet the statutory requirements.
ACIS believes that the primary statute for controlling the abuse of intellectual property rights that are or become de facto NII interface standards should be the intellectual property law, not the antitrust laws.[3] This however does not mean that antitrust laws can abdicate any responsibility of preventing monopolistic abuse. Recent trends in antitrust theory suggest that the antitrust laws will exercise little restraint on monopoly power deriving from intellectual property claims in NII standards. The antitrust enforcement agencies should closely monitor the evolution of the NII and the emergence of de facto standards. They should be prepared to intervene if any vendor appears to misuse its control of a de facto standard. Further, they should approve mergers and joint ventures relating to the NII only if the parties involved in the transaction waive any proprietary claim to de facto NII interface specification standards.
SUN Microsystems believes that critical NII software and hardware interface specifications should be neither patentable nor copyrightable. Further, the interface specifications should be published, eliminating the need for reverse engineering.
The approach proposed by this camp appears to balance vendor interests and consumer interests very well. Some issues still need to be resolved here though. Most of the organizations in this camp say only the "critical interfaces" should be open. It is not clear how one would go about identifying the critical interface. One of the reasons for the hot competition in the computer industry was the desire to gain market acceptance of a protected interface standard. If interface standards are not protected anymore the pace of development of the National Information Infrastructure may slow down.
This camp believes that the U.S. should not interfere with the National Information Infrastructure. They believe that not only should free market choose the interface standard, but also that these interface standards should be protectable as intellectual property. Organizations subscribing to this view include Business Software Associates (BSA) whose members are companies like Microsoft, IBM, Lotus and Novell. We examine below Microsoft's arguments in support of this approach.
Microsoft believes that buyers of communication and information systems value interoperability, but it is only one of the many competing values along with price, speed, flexibility, ease-of-use, etc. Ultimately, the marketplace is the most efficient determinant of which features and benefits are important, and at what price. In computer industry, interoperability successfully evolved in the open market driven by competition and innovation. In addition, the open market process allowed technology to continue to evolve, rather than freezing it at obsolete levels. Recent history has proven markets to be the fastest and most efficient mechanism for defining interoperability and its implementation. Standards-setting organizations can and should agree on interoperable interfaces as they are validated by the marketplace.
Microsoft believes that interoperability has been oversold to people. There are costs entailed by interoperability. Designing a system that interoperates with a multitude of other systems may be useful to some customers, but useless to others. Naval air defense systems and household appliance controllers may both be advanced information systems but there is probably little need for them to interoperate. Interoperability is a customer consideration or value in acquiring an advanced information system, but it may not be the customer's primary value. Speed, security, flexibility, innovation, cost effectiveness, ease-of-use, and other values may be more important to some customers. Providing interoperability on the scale envisioned may be very costly. Should advanced information systems be required to interoperate with Apple IIs, Commodore 64s, IBM360s? 100% interoperability of all systems may not be the best goal. 80% interoperability may probably be better from a cost benefit trade-off perspective. Achieving the rest 20% interoperability might double the costs.
Perhaps Microsoft's most vocal objections(and worst fears) are the de jure standards. De jure means "according to law; by right."[7] These standards are official or legal standards established by government or government bodies. The Federal Communication Commission's 1953 standard for broadcasting color television, known as NTSC(National Television Systems Committee), was a de jure standard. Broadcasters and television manufactures were required to adhere to NTSC standards in order to operate within the United States. Further, the standards set by the government based on projections or expectations of future developments are called anticipatory standards.
Microsoft believes that anticipatory de jure standards are harmful to the growth of the National Information Infrastructure. Companies will invest less on developing a standard and more on lobbying to get their standard approved. Microsoft points out that the computer revolution was successful without the government interfering. If the government had tried to set a PC standard in the early eighties we would probably still be using Apple IIs, or Radio Shack TRS-80. The other danger is setting of a premature standard. Academics, industrialists, and politicians have been talking about the information revolution for decades. It is difficult to know what stage of this revolution we are in. Are we at the beginning or at the end? It is difficult to answer. Locking in a standard too early might stifle innovation and slow down progress. Thomas J. Watson, long time Chairman of Board at IBM once said "I think there is a world market for about five computers". Microsoft also points out that government has a poor record of standard setting. The efforts by Japanese and European governments to set HDTV standards, accelerate HDTV research and gain official approval of anticipatory standards for HDTV is exemplary. Both the governments set analog anticipatory standards. The Japanese standard was set in the 1970s and the European in the 1980s. As we moved into the 1990s it was clear that digital technology was more suited for HDTV. U.S., not bound by any analog standards is now pioneering the HDTV effort. Japanese and Europeans not only lost billions of dollars but also the lead in HDTV research.
Microsoft also supports protection of interface standards. They believe that non proprietary is the information world's version of the proverbial "free lunch." Microsoft believes that non proprietary interfaces do not provide adequate incentive for innovation. They will not enable U.S. firms to withstand foreign competition. Vendors are often motivated to solve problems in proprietary systems than in open systems. Customers tend to be happier because their vendors are serious about solving problems. Microsoft fears that if standards are non proprietary they will not evolve, government will take over the job and refer it to one of the committees which will take us back to where we started -- de jure anticipatory standards.
Microsoft conveniently omits a few things from its analysis. It neglects to mention that protection of network operating system interface standards would give Microsoft control of a billion dollar market, leading to profits orders of magnitude larger than its R&D investment. Also, foreign firms are getting technologically more sophisticated and if they manage to set the standard all U.S. firms will be captive to their standard.[6] Microsoft's arguments also misrepresent the position of most interoperability advocates. Most interoperability supporters want interface specifications to be non proprietary. Microsoft is trying to divert the attention of the whole debate by making it sound that all its opponents want a de jure standard. Microsoft wants to convince everyone that nonprotectable standards are not viable and that eventually it would force the government to set standards. There are other organizations like Computers Systems Policy Project(CSPP) which subscribe to Microsoft's view except that they advocate that standard owners should be bound by voluntary licensing policy which is reasonable and non discriminatory. This argument, modeled on the patent policy of most standard organizations, has a fundamental flaw in the NII context. The standard organizations can enforce this policy in case of patents. In case of NII there is no single standards organization and hence a lack of enforcement mechanism.
There are definitely valid arguments for the three positions. Presence of network externalities is an important argument for some amount of government regulation. Left to market forces, telephone would probably have not been a universal service now. However it took more than fifty years for everyone to be connected. Today technology is moving faster than ever. The fat slow bureaucratic nature of the government agencies is not well suited for recommendation of anticipatory standards. It is best for the government to keep out of the realm of standard recommendation. However the government has two roles to play in the building of the National Information Infrastructure. First, formulating appropriate law to prevent market domination by few firms, and second, ensuring universal service to the society.
The Maximal protection approach appears to be dangerous too. It is excessively tilted in favor of a few companies who want the whole pie. Microsoft argues that it is only fair that companies that invest in interfaces should be able to benefit from it. Some of the most critical interface specifications became standards not because companies spent a large amount of money and time to come up with the best specifications, but rather because the interface was bundled together with a system or application that was a compelling product due to technological or market reasons. MS-DOS was not a good operating system even in by the early eighties standard- the fact that Tim Patterson who wrote the initial version called it Quick and Dirty Operating System(QDOS) should be a hint. What made it popular was the IBM PC and applications like Lotus 123. Since operating system is the gateway between hardware and applications, a copyright on MS-DOS has given Microsoft enormous control over the entire Personal Computer industry. Rivals want the "application programming interface" placed under industry wide control. Microsoft refuses, saying it is playing tough but fair. The serendipitous path to interface standards does not justify rewarding investment in interface specifications with the enormous leverage afforded by protection of interface standards. The other danger of giving monopolies ownership of interface standards is the possible suppression of some fledgling technologies. Monopolies will try to guide the evolution of NII towards exploiting their core competencies which will entrench existing technologies and prevent new paradigms from emerging.
The Open Interface approach advocated by ACIS appears to be the most reasonable solution. Non proprietary interface standards will encourage healthy and fair competition for products on either side of the interface. There have been arguments that non proprietary interface standards will not encourage the industry to invest in interface specifications. This is a fallacious argument because there is a huge monetary benefit to be gained by being first to market with certain key applications that will need to communicate with other applications across NII interfaces. The benefits are incentive enough for companies to define interface specifications.
On May 24 1995 Microsoft and BSA won a victory as the House commerce committee adopted an amendment by Rick White (Republican- Washington) to the telecommunications bill that would let the market determine the operating systems for television set-top boxes. The original bill language would have given the Federal Communications Commission authority to order an open operating system and interoperability. In a letter to the committee members, White argued "Government intervention in setting standards for the way information is created, processed and used by computers would likely second-guess private industry and the marketplace, undermine intellectual property rights inhibit innovation, and reduce competition."[8]
It is unfortunate that the House commerce committee did not understand the real issues and/or gave into pressure from Microsoft and BSA. The ACIS consortium needs to exercise more clout if they are to prevent Microsoft and BSA from bulldozing their way through this issue.
[1]In pursuit of an Optimum: A conceptual model for examining public sector policy support interoperability, Lori Annette Perine, National Institute of Standards and Technology. July 1995 draft
[2]Perspectives on NII Interoperability: An Industry strategy for Improving the Interoperability of Information-Rich Documents, The Computer System Policy Project, Draft, May 1995
[3]Intellectual property and the National Information Infrastructure - ACIS comments
[4]Comments on International Aspects of the National Information Infrastructure. American Committee for Interoperable Systems
[5]Standards: The Rough Road to the Common Byte Martin C. Libicki, Standards Policy for Information Infrastructure, MIT Press.
[6]Competing Definitions of "Openness" on the NII - Jonathan Band, Standards Policy for Information Infrastructure, MIT Press.
[7]Interoperability and Advanced Information Systems: The Problems of Anticipatory and De-Jure-Official Standards. Microsoft White paper presented at the Interoperability and the Economics of Information Infrastructure. workshop, July 6-7, 1995.
[8]Microsoft, BSA Win in Telecom Bill Markup 05/25/95, Newsbytes News Network, May 25, 1995
[9]Midrange Systems, June 30, 1995
[10]Standards Cowboys Vs. Committees, Peter Coy, Business Week, April 10, 1995
[11] Arguments for Weaker Intellectual Property Protection in Network Industries, Joseph Farrell, Standards Policy for Information Infrastructure, MIT Press.
[12]Statement of Cita Furlani Program Manager, Enterprise Integration, Computer Systems Laboratory, NIST, before the Subcommittee on Technology, Environment, and Aviation and the Subcommittee on Science, Committee on Science, Space, and Technology, House of Representatives, May 26, 1994, on electronic commerce and interoperability in the NII .